1

Helping Caregivers and Their Children With Early Appendicitis Make Treatment Decisions With an App

Katherine Deans, MD, MHSc1, Peter Minneci, MD, MHSc1, Lai Wei, Ph.D. 2, Kyle Porter, M.A.S. 2,

Kelly Kelleher, MD, MPH1

1Research Institute at Nationwide Children’s Hospital, Columbus, OH 2The Ohio State College of Medicine Department of Biomedical Informatics, Columbus, OH

Original Project Title: Randomized Controlled Trial of a Patient Activation Tool in Pediatric Appendicitis PCORI ID: CE-12-11-4350 HSRProj ID: 20143580 ClinicalTrials.gov ID: NCT02110485

_______________________________ To cite this document, please use: Deans K, et al. (2019). Helping Caregivers and Their Children With Early Appendicitis Make Treatment Decisions With an App. Washington, DC: Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/02.2020.CE.12114350

2

Table of Contents ABSTRACT .................................................................................................................. 4

BACKGROUND ............................................................................................................ 6

PARTICIPATION OF PATIENTS AND OTHER STAKEHOLDERS ....................................... 10

Balance of Stakeholder Perspectives ............................................................................ 10

Methods of Stakeholder Identification and Recruitment ............................................. 10

Types and Numbers of Stakeholders ............................................................................ 10

Stakeholder Engagement and Impact During First 6 Months ....................................... 11

Stakeholder Engagement and Impact After 6 Months ................................................. 14

Table 1. Summary of Stakeholder Contributions ................................................................ 15

METHODS ................................................................................................................ 17

Study Design .................................................................................................................. 17

Forming the Study Cohort ............................................................................................. 19

Table 2. Inclusion/Exclusion Criteria ................................................................................. 20

Study Setting ................................................................................................................. 21

Interventions ................................................................................................................. 21

Table 3. Comparison of Content Presented in the Standardized Surgical Consultation and PAT .................................................................................................................................. 22

Follow-up ....................................................................................................................... 23

Study Outcomes ............................................................................................................ 23

Primary Outcomes ......................................................................................................... 23

Table 4. Study Outcomes .................................................................................................. 24

Secondary Outcomes .................................................................................................... 27

Data Collection and Sources ......................................................................................... 30

Figure 1. Study design....................................................................................................... 31

Analytical and Statistical Approach ............................................................................... 32

Conduct of the Study ..................................................................................................... 36

RESULTS ................................................................................................................... 37

Enrollment and Follow-up ............................................................................................. 37

Table 5. Reasons for Exclusion of 914 From the Studya ...................................................... 38

Figure 2. CONSORT flow diagram ...................................................................................... 39

Table 6. Follow-up Rates by Randomized Group and Treatment Choice ............................. 41

3

Table 7. Baseline Demographics and Clinical Characteristics for the 2 Groups .................... 42

Interim Analyses ............................................................................................................ 43

Primary Outcomes ......................................................................................................... 43

Table 8. Primary Outcomes ............................................................................................... 45

Figure 3. Distribution of scores on the Decisional Self-Efficacy Scale at 1 hour .................... 46

Decision Making and Patient-Centered Outcomes ....................................................... 46

Table 9. Decision Making and Patient-Reported Outcomes ................................................ 47

Medical Outcomes ........................................................................................................ 48

Table 10. Medical Outcomes at 30 Days, 6 Months, and 1 Year .......................................... 49

Effect of Treatment Choice on Primary Outcomes Based on Randomized Group (HTE) . ....................................................................................................................................... 51 Table 11. Primary Outcomes by Randomized Group and Treatment Choice ........................ 53

DISCUSSION ............................................................................................................. 54

Decisional Context ......................................................................................................... 54

Study Results in Context................................................................................................ 54

Implementation of Study Results .................................................................................. 55

Generalizability .............................................................................................................. 55

Subpopulation Consideration ....................................................................................... 56

Study Limitations ........................................................................................................... 56

Future Research ............................................................................................................ 57

CONCLUSIONS .......................................................................................................... 58

REFERENCES ............................................................................................................. 59

PUBLICATIONS ......................................................................................................... 67

APPENDIX ................................................................................................................ 70

4

ABSTRACT Background: Patient activation is the process of establishing a patient’s willingness to take an active role in his or her care, ensuring a patient’s confidence in his or her ability to manage his or her care, and providing a patient with the knowledge and skills necessary to make decisions and manage his or her own care. Previous adult studies have demonstrated that patient activation can improve healthy behaviors, increase the appropriate use of health care resources, improve patient preparedness for medical appointments, and improve the management and control of chronic illnesses such as diabetes. However, few studies have examined the effects of patient activation on decision making and patient-centered outcomes in children, the acute care setting, or surgery. Emergency surgical interventions present stressful and difficult decisions for patients and caregivers. This is especially true in vulnerable populations such as children, in which caregivers are forced to take in large amounts of information and make quick decisions for someone other than themselves. Strategies to activate caregivers in the emergency setting through knowledge and engagement may result in improved decision making with improved self-efficacy, confidence, and satisfaction.

Objectives: To determine if a patient activation tool (PAT) can activate pediatric patient–caregiver dyads to make confident, informed decisions related to the treatment of appendicitis, an acute surgical disease.

Methods: We performed a prospective, single-institution, single-blind, randomized controlled trial comparing an electronic and interactive PAT to standardized consultation. Our study population included 200 children aged 7-17 years with acute, uncomplicated appendicitis and their caregivers and ran from March 2014 through April 2016. Patients and their caregivers were randomized to either a scripted standardized surgical consultation alone or to a scripted standardized surgical consultation plus the PAT. A physician member of the research team administered both interventions. The scripted standardized consultation represented the best available standard of care and included an explanation of the risks and benefits of each treatment choice and of the importance of patient–caregiver preferences and values when making this treatment decision. The PAT is a tablet-based tool designed to present the risks and benefits of each treatment choice, activate the patient–caregiver dyad to participate in medical decision making, and alleviate decisional uncertainty through a variety of media. Patients and caregivers were asked to choose either nonoperative management with antibiotics alone or surgery for the treatment of appendicitis. We followed all patients for 1 year. The primary outcomes were decision self-efficacy, health care satisfaction, and disability days. Specifically, we tested for superiority of the PAT compared with standardized surgical consultation on decisional self-efficacy and health care satisfaction and for noninferiority of the PAT on days of disability. Secondary outcomes included knowledge, decisional conflict, health-related quality of life, length of hospital stay, rate of readmission, and recurrence of appendicitis. We performed analyses considering each patient–caregiver dyad as a unit, using the caregiver’s responses to all surveys. We compared dichotomous outcomes using Fisher exact tests and continuous outcomes using Wilcoxon rank sum tests. We performed subgroup comparisons

5

based on the randomized intervention and treatment choice using Kruskal-Wallis tests for continuous variables and Fisher exact tests for categorical variables.

Results: Ninety-eight patients were randomized to the PAT and 102 to the standardized consultation group. Decisional self-efficacy within 1 hour of making a treatment decision was higher in the standardized consultation group compared with the PAT group (median [interquartile range]: 100 [97.7-100] vs 100 [95.5-100]; p = 0.03). The percentage of patients who chose each treatment was similar in both groups (standardized consultation: 41% antibiotics and 59% surgery; PAT: 32% antibiotics and 68% surgery; p = 0.19). Total scores on health care satisfaction at discharge were similar between groups (99 [94.7-100] vs 98 [91.7-100]; p = 0.27). Disability days at 1 year were also similar (6 [2-11] vs 5 [2-15]; p = 0.67). We found no differences in rates of hospital readmission at 1 year (16.3% vs 18.1%; p = 0.84) or emergency department/urgent care visits at 1 year (6.6% vs7.2%; p = 1.00). Among those patients who underwent surgery (primary choice plus failure of nonoperative management), we found no differences between the standardized consultation and PAT groups in rates of complicated appendicitis (10.3% vs 12.7%; p = 0.79). Among those patients who chose nonoperative management we found no difference in the treatment failure rate of nonoperative management at 1 year between standardized consultation and PAT groups (34.2% vs 36.7%; p = 1.00) nor in the rate of complicated appendicitis (5.3% vs 3.3%, p = 1.00).

Conclusions: A technology-based tool did not improve measures of decision making or patient activation for patients and families needing to make a medical treatment decision in an emergency care setting compared with a standardized consultation that emphasized patient choice. The very high median scores for decisional self-efficacy in each group indicate extremely high self-efficacy and the detected statistical differences may not be clinically meaningful. These results may reflect a ceiling effect of the survey instrument or that an intervention to improve decision making in this setting may not be needed.

Limitations: The PAT intervention did not improve our primary outcomes compared with standardized surgical consultation at our institution. This might have been in part due to the standardized consultation in the control group, which we designed to represent the best available standard of care, not usual care. In addition, the same physician staff members who administered the PAT administered the standardized consultation , which might have led to contamination. As such, the control group may not be an accurate reflection of usual practice across heterogeneous settings.

6

BACKGROUND According to the US Department of Defense Patient Safety Program, there are 3

essential components of patient activation: (1) establishing patients’ willingness to take an

active role in their care, (2) ensuring patients’ confidence in their ability to manage their care,

and (3) providing patients with the knowledge and skills necessary to make decisions and

manage their own care.1 In adult patients, patient activation has been shown to increase the

appropriate use of health care resources and to improve healthy behaviors, patient

preparedness for medical appointments, and the management and control of chronic illnesses

such as diabetes.2-7 There is little evidence examining the impact of patient activation on

decision making and patient-centered outcomes in the emergency care setting or in pediatric

surgery.

Emergency surgical interventions present stressful and difficult decisions for patients

and caregivers. Surgeons provide generalized information about treatment options to acutely ill

patients with limited time for shared decision making. This is especially true in vulnerable

populations such as children, in which caregivers are forced to take in large amounts of

information and make quick decisions for someone other than themselves. We hypothesized

that strategies to activate caregivers in the emergency setting, through knowledge and

engagement, might result in improved decision making with improved self-efficacy and

confidence. This could lead to improvements in other patient-centered outcomes, such as

health-related quality of life (HRQOL) and health care satisfaction, for pediatric patients as well

as their caregivers. To test these hypotheses, we used acute appendicitis and the treatment

decision faced by patients and families between (1) surgery and (2) nonoperative management

with antibiotics alone.

Approximately 136 million Americans, including more than 3.4 million people with

abdominal pain, seek emergency medical care each year.8,9 Appendectomy (surgical removal of

the appendix) is the most common emergency operation performed for abdominal pain in both

adults and children.9 More than 250 000 appendectomies are performed in the United States

7

each year, of which 70 000 are performed in children.10,11 Although surgery represents a

curative treatment for appendicitis, it is associated with a period of pain, disability, and the

potential risk for a surgical complication. Appendectomy is a major intra-abdominal procedure

that requires general anesthesia. Perioperative complications occur in 8.7% to 11.1% of

patients, and recovery from an uncomplicated appendectomy requires children to miss 1 to 2

weeks of school and their caregiver to miss equal amounts of work.12,13 Nonoperative

management of appendicitis with antibiotics alone has recently emerged as a viable treatment

alternative to surgery and is being increasingly used.12, 14-38 Nonoperative management with

antibiotics alone can successfully treat 3 out of 4 patients with less disability and pain while

avoiding the potential risks of an operation. However, 1 out of 4 patients will eventually need

an appendectomy. These 2 treatments have very different risk–benefit profiles and the best

choice for an individual patient depends on which risks and benefits are most important to each

patient and his or her family. For example, families that live in remote areas or that are very

fearful of appendicitis recurrence may more frequently present to emergency departments

when their child has abdominal pain. This may make surgery appear more beneficial. On the

other hand, nonoperative management with antibiotics alone may be the preferred choice for

families that are averse to surgery, the need for general anesthesia, and the inherent risks of

undergoing an operation.

Within this study we attempted to address 3 specific evidence gaps related to shared

decision making in pediatric surgical patients. First, most studies on patient activation, decision

making, and patient-centered outcomes focus on autonomous adult patients.39-43 Few studies

have examined patient–caregiver dyads in the context of clinical decision making in children.44-

50 This study specifically examined patient and caregiver activation in pediatric patient–

caregiver dyads. Second, most research on patient activation, decision making, and patient-

centered outcomes is focused on chronic illness and/or outpatient care in which there is a

significant amount of time (days to weeks) for patients to reflect on information and practice

communication strategies.39-43 However, high acuity illness is directly related to patient and

caregiver stress, decreased comprehension, and impaired recall.51-57 This study specifically

8

examined patient activation in the emergency care setting. Third, studies investigating

components of patient activation in surgical settings have been performed only in adult

patients undergoing major elective surgery, with none targeting emergency surgical procedures

or children and their caregivers.43 This study investigated the effects of patient and caregiver

activation on decision making and patient-centered outcomes in an acute surgical disease in

children. There were no systematic reviews to highlight these evidence gaps; therefore, we

performed our own comprehensive review of the literature.

Our goal was to determine if a patient activation tool (PAT) can activate pediatric

patient–caregiver dyads to make more confident, informed decisions about a treatment course

for appendicitis. The definition of activation we used in this project stemmed from a

combination of the conceptual model proposed by Hibbard et al and the Ottawa Decision

Support Framework.58,59 In this study, we defined an activated patient–caregiver dyad primarily

on the willingness, knowledge, engagement, and self-efficacy of the caregiver. Consequently,

we developed an integrated PAT that activates the patient–caregiver dyad, provides knowledge

and skills regarding the medical decision, and strengthens self-efficacy in order to assist the

caregiver in his or her treatment choice. We attempted to demonstrate that a PAT can be

successfully used in an emergency care situation to improve decision making and patient-

centered outcomes (activation, knowledge, preparedness of decision making, decision self-

efficacy and regret, decisional conflict, satisfaction, and HRQOL) without compromising medical

outcomes (disability days, hospital readmission rates, and hospital length of stay).

The study’s aims and hypotheses were the following:

Specific Aim 1: To perform a randomized controlled trial (RCT) comparing a PAT to

standardized surgical consultation in patient–caregiver dyads choosing between nonoperative

management and appendectomy for early appendicitis in order to identify differences in

measures of decision making and patient-centered outcomes. Hypothesis: A PAT that activates

patient–caregiver dyads will improve decision making and patient-centered outcomes such as

caregiver decision self-efficacy, preparedness for decision making, decisional conflict, decision

9

regret, caregiver activation, caregiver and child satisfaction, HRQOL, and caregiver knowledge.

This aim tests the superiority of the PAT compared with standardized surgical consultation.

Specific Aim 2: To characterize the effects of a PAT on medical outcomes of appendicitis

in patients receiving the PAT compared with those receiving standardized surgical consultation

alone. Hypothesis: A PAT that activates patient–caregiver dyads will not compromise medical

outcomes including disability days, hospital length of stay, hospital readmission rates, and

medical complications related to treatment choice (eg, infection, recurrence). This aim tests the

noninferiority of the PAT compared with standardized surgical consultation.

To our knowledge, this was the first study to combine a patient activation strategy with

a decision aid for patient–caregiver dyads in an emergency surgical setting. Results from this

study could be informative about the utility of patient activation in this setting. A positive result

from our study could lead to the rapid development and implementation of PATs for

appendicitis at most children’s hospitals in the United States and Canada. In addition, because

the underlying structure of the PAT is modular, it can be adapted and studied for use in other

emergency and urgent surgical procedures in both children and adults.

10

PARTICIPATION OF PATIENTS AND OTHER STAKEHOLDERS Patient, family, and other health care stakeholders were an integral part of the research

team throughout the entire project. Their participation and commitment were critical to the

successful completion of this project. Our multidisciplinary stakeholder team provided unique

perspectives that helped us formulate the research question, design the intervention and the

study, improve recruitment and retention throughout the study, and plan for dissemination.

Balance of Stakeholder Perspectives

To facilitate designing a study with results that would be meaningful to as many end

users as possible, the research team believed that our stakeholder group should incorporate

representatives involved in all phases of pediatric appendicitis care. This included patients with

appendicitis and their families, nurses, pediatricians, emergency medicine physicians, surgeons,

patient educators, and health care payers.

Methods of Stakeholder Identification and Recruitment

We identified stakeholders through a variety of mechanisms. including an Institutional

Review Board (IRB)-approved flyer distributed to the Nationwide Children’s Hospital (NCH)

community, face-to-face recruitment through the NCH outpatient surgical center and within

the institution’s Family Advocacy Council, and personal communication and solicitation by the

investigators.

Types and Numbers of Stakeholders

Our stakeholder group comprised 6 patient-family dyads and 5 other health care

providers, including a surgery clinic nurse, a child–family educator, an emergency medicine

physician, a community pediatrician, and a payer representative. They participated in individual

interviews, group meetings, and email/phone communications. Our stakeholders helped us

develop the PAT, design the study, monitor and improve our patient enrollment and follow-up,

and plan for dissemination of our results. We sought active input from all of the stakeholders

11

throughout the study during the in-person interviews, group meetings, and the email/phone

communications with all opinions and perspectives valued equally. The in-person group

meetings provided a forum to openly discuss issues and reach consensus among all

participants. Minutes from each meeting were recorded, distributed to the group, and

referenced if needed for clarification. Table 1 at the end of this section describes the

contributions of our stakeholders throughout this project.

Stakeholder Engagement and Impact During First 6 Months

We integrated stakeholders into the study team from the study’s inception. Initial

stakeholder involvement, during the first 6 months of this project, included several rounds of

individual interviews, in-person group meetings, phone calls, and email communications to

develop the research question, design the study, develop the intervention, and choose the

outcomes for this project. We audio-recorded individual interviews with stakeholders. A

member of the research team reviewed each recording. We compiled, grouped, and

summarized the comments based on content. We deidentified the comments and presented

them to all stakeholders for discussion.

During the initial patient and family stakeholder interview (which occurred prior to

receiving PCORI funding), we asked families to describe the most challenging aspects of their

appendicitis care. They unanimously agreed that making the decision between an

appendectomy or antibiotics alone was the most difficult part of the process. Providers also

agreed it was difficult to take the necessary time to fully explain the risks and benefits of each

choice within the time constraints of an encounter in the emergency department. They also

described that not all families seemed to be interested in making a choice or understanding

that they have a choice. Consensus from these interviews suggested that both families and

providers felt that it was important to further study methods for improved shared decision

making. After completing the initial interviews, we held a group meeting to review the

feedback, the proposed study question, and the study design. After lengthy deliberation, the

group decided it would be worth studying a tablet-based tool that could describe the treatment

12

choices and provide assistance in aligning each family’s values and preferences with a

treatment choice. Based on the feedback from the providers, we agreed that a patient

activation module should be incorporated to facilitate engagement. The research team then

proposed an RCT that compared a standardized consultation to a PAT that was designed to

engage, educate, and activate patients and their caregivers in a tablet-based application.

Subsequently, we selected the primary and secondary patient-centered outcomes in this study

with the assistance of our stakeholders. We believe this early and intensive collaboration with

our stakeholder group helped facilitate a committed partnership throughout the course of the

study.

We then began an iterative process with the stakeholders to develop the PAT. In

addition, we collaborated with members of the Ottawa Patient Decision Aids Research Group to

ensure the decision aid component of the PAT was consistent with the Ottawa Decision Support

Framework and International Patient Decision Aid Standards (IPDAS). We developed lists of the

potential risks and benefits of each treatment option. We reviewed these lists with the

stakeholders by email to ensure completeness and comprehension. We created paper versions

of these lists and developed scripts that explained the 2 treatment options and their risks and

benefits. At our second in-person group meeting, we reviewed this content and had an open

discussion about the format for the PAT. Potential options raised included paper, audio, video,

and interactive modalities using a computer or tablet. Consensus was that an interactive tablet-

based PAT would be the most engaging to both children and caregivers. We also decided that

video vignettes portraying patients and caregivers might be helpful to the patients and

caregivers faced with making the same treatment decision.

The research team then worked with a film company and a software company to

develop a prototype PAT that could be viewed and used by both the patients and their

caregivers. We reviewed the prototype with each stakeholder individually prior to our next

group meeting, wherein we reviewed the feedback and came to consensus for additional

changes. Stakeholder feedback during this stage provided (1) additional information on the

importance of developmental stage, age, and learning style when discussing treatment options

13

with children and (2) information on the most effective methods for engaging and teaching

patients and their caregivers about their illness and treatments. More specifically, stakeholder

feedback in the design of our intervention led to (1) the inclusion of multiple different avatars

that were sensitive to skin color, gender, and age; (2) a time limitation of <20 minutes for the

length of the experience; (3) additional vignettes from patients and caregivers; (4) a specific

visual format for how the treatments were described; (5) the format of our 2 x 2 grid comparing

the risks and benefits of each treatment; (6) graphics that assisted with low numeracy levels;

(7) the specifics of an interactive values exercise that helps patients prioritize the risks and

benefits for themselves; and (8) an optional module that provides additional training on

techniques for communicating with health care providers. The final version of the PAT met 18

out of 23 content criteria, 20 out of 23 development criteria, and 7 out of 7 effectiveness

criteria established by IPDAS. We reviewed the trial logistics during another in-person group

meeting in which the stakeholders strongly suggested that we revise our enrollment process to

have a physician perform the enrollment and administer the assigned intervention rather than

a research nurse (as described in the original proposal). The stakeholders believed patients and

families in the acute care setting would be more agreeable to hearing about research and

different treatment options directly from a physician who could both perform the study-related

procedures and answer questions. Based on this feedback, we changed our enrollment process

to use physicians on the research team to perform enrollment and administer the assigned

intervention. In addition, the stakeholders believed the control group should receive a

standardized consultation that represented the best available standard of care and not usual

care. Consensus was that, in addition to explaining the risks and benefits of each treatment

choice, the standardized consultation should also include an explanation of the importance of

patient–caregiver preferences and values when making this treatment decision. We developed

and reviewed the standardized consultation script with our stakeholders and incorporated their

feedback.

14

Stakeholder Engagement and Impact After 6 Months

After the first 6 months of the study, we engaged stakeholders at least 4 times per year

with semiannual, in-person group meetings and quarterly phone/email communications.

Following each semiannual meeting, we sent study updates to communicate the final actions to

be implemented on issues discussed during the meeting. We used quarterly email/phone

updates to (1) inform stakeholders on the progress of the study, (2) communicate any results

that emerged due to stakeholder input, and (3) solicit questions and suggestions to increase the

study’s success. These methods led to the development of strong working relationships, which

have facilitated the participation of each member of the stakeholder team throughout the

study.

One striking example of how stakeholder input affected this study was on participant

recruitment and retention.60 Our initial recruitment rate was 65%. We discussed this at one of

our in-person stakeholder meetings and asked for suggestions on how we might improve our

current recruitment process. Our stakeholder group asked us to “role play” our recruitment and

consent processes. Based on this, stakeholders made suggestions on how to change our

enrollment script from one that stated “We are investigating a tool designed to improve

decision making about appendicitis treatments” to a script with a 2-part message. The first part

stated that we worked with patients and families to design this study to improve the way the

medical team communicates with families. The second part explained that this study was

testing a tool designed to improve physician–patient communication and to promote shared

decision making about treatments for their child’s appendicitis. After making this change, trial

enrollment significantly increased, from 65% to 95%.60 We also discussed our retention rate

with the group. Our initial retention rate was 58%. We reviewed our follow-up processes with

the stakeholders and asked for suggestions. Based on stakeholder recommendations, we began

incorporating an email/online option for participants to complete follow-up questionnaires.

This increased the rate of completed 30-day follow-up from 58% to 85%.60 We maintained both

of these gains throughout the duration of the study and ended up with an 86% enrollment rate

and an 87% 1-year follow-up rate.

15

Table 1. Summary of Stakeholder Contributions

Defining the Decisional Context

Identified that decision making between an appendectomy or antibiotics alone was the most difficult part of the process

Recognized that it is difficult to take the necessary time to fully explain the risks and benefits of each choice within the time constraints of an encounter in the emergency department

Determined that it was important to further study methods for improving shared decision making

Developing Study Question and Design

Selected randomized controlled design to compare a standardized consultation with a patient activation tool (PAT)

Determined that the goals of the intervention should be to engage, educate, and activate patients and their caregivers

Favored a tablet-based application as the intervention

Suggested inclusion of a patient activation module to facilitate engagement

Selected the primary and secondary patient-centered outcomes

Development of the PAT

Decided that an interactive tablet-based PAT would be the most engaging to both children and caregivers

Selected lists of the potential risks and benefits of each treatment option

Specified content based on consensus of stakeholder input:

Inclusion of multiple different avatars that were sensitive to skin color, gender, and age

Time limitation of < 20 minutes for the length of the experience

Additional vignettes from patients and caregivers

Specific visual format for how the treatments were described

2 x 2 grid format comparing the risks and benefits of each treatment

Graphics that assisted with low numeracy levels

16

Table 1. Summary of Stakeholder Contributions

Specifics of the interactive values exercise that helps patients and caregivers prioritize the risks and benefits for themselves

Inclusion of an optional module that provides additional training on techniques for communicating with health care providers

Development of the Standardized Consultation

Suggested that it should represent best available standard of care Recommended that it minimize variation in how treatment options are explained across surgeons

Specific content based on consensus of stakeholder input:

Explanation of the risks and benefits of each treatment choice

Explanation of the importance of patient–caregiver preferences and values when making medical decisions

Trial Logistics

Revised enrollment process to have a physician perform the enrollment

Changed recruitment script

Changed follow-up process

17

METHODS

Study Design

The choice between appendectomy and nonoperative management for acute

appendicitis must balance the unique risk–benefit profile of each treatment with the

preferences and values of each individual patient–caregiver dyad. Determining the best

treatment requires engaging the patient and caregiver in the treatment decision. We

conducted an RCT to determine if providing patient–caregiver dyads with a PAT improves

decision making and patient-centered outcomes without compromising medical outcomes

compared with standardized surgical consultation. Based on the essential components of

patient activation, which include establishing a willingness to take an active role in their care,

ensuring confidence in their ability to manage their care, and providing the knowledge and skills

necessary to make a treatment decision, we designed the PAT as an interactive electronic tool

that integrated both a decision aid and patient–caregiver activation and communication

strategies. The goal of the PAT was to activate the patient–caregiver dyad, provide knowledge

and skills regarding the medical decision, and strengthen self-efficacy in order to assist in the

treatment choice. Because both the PAT and standardized surgical consultation were provided

during a brief period prior to the treatment decision, we expected a negligible lack of

adherence in either group. We used the randomized controlled design to minimize selection

bias and allow for causal estimates of the effect of a PAT on our chosen outcomes.

This study randomized pediatric patients with appendicitis and their caregivers to either

a scripted standardized surgical consultation alone or to a scripted standardized surgical

consultation plus the PAT. Both randomized groups were given access to the same information.

One group received the information through an interactive, tablet-based PAT and the other

group received information through a scripted standardized surgical consultation. Both groups

received this information prior to meeting the attending surgeon. The direct comparison

between the PAT and the standardized surgical consultation alone allowed us to causally

determine if the addition of a PAT improved measures of decision making and patient-centered

18

outcomes in pediatric patient–caregiver dyads. Compared with the scripted surgical

consultation, the PAT provided additional content such as video explanations of the treatments

and their risks and benefits, vignettes of caregivers and children explaining the most commons

reasons for choosing each treatment, an interactive exercise to help caregivers and patients

align their preferences with their treatment choice, and embedded activation and

communication strategies. Because we designed the PAT as an interactive tablet-based

application, it allowed the user to move at his or her own pace and review the content as

needed. The PAT is a hybrid intervention designed to increase the patient–caregiver dyads’

willingness to take an active role in their care, improve their confidence in their ability to

manage their care, and provide them with the knowledge and skills necessary to make

decisions and manage their own care.

Working with our co-investigators, we developed the PAT based on the concept of the

patient activation continuum and the potential for improved outcomes in more activated

patients compared with less activated patients.1 Previous research has suggested that teaching

patients to effectively communicate with clinicians leads to increased patient participation in

their health care, improves a variety of patient communication variables, positively effects

important health outcomes, and changes the nature of the physician–patient interaction.42,61 In

our study, the PAT had embedded communication and engagement strategies based on the

components of the PACE (Presenting, Asking, Checking, Expressing) system, which was

developed by a member of our research team (DM Post) and adapted for use in this study.62-65

The PACE system is a patient education system designed to improve communication with

physicians by emphasizing the acquisition of proficiency in 4 specific categories: (1) presenting

detailed information; (2) asking questions if desired information has not been provided; (3)

checking understanding of information that is given; and (4) expressing any concerns about the

recommended treatment. We designed the script, vignettes, illustrations, and exercises

contained in the PAT to present patients and their caregivers with detailed and understandable

information about the 2 treatment options, encourage them to ask questions about the

19

treatments, check their understanding of the treatment options, and encourage them to

express concerns about each treatment option specific to their child.

We hypothesized that the PAT would help patients and their caregivers communicate

more effectively with the medical team and choose the treatments that were best aligned with

their preferences, values, and family circumstances. This would directly improve decision

making and patient-centered outcomes and indirectly could improve disability days. For

example, for families that do not want to accept the risk of recurrent appendicitis,

nonoperative management can lead to more disability. If a family is so afraid of a recurrence

that they bring their child to the emergency department every time he or she has abdominal

pain, then their child will likely undergo increased imaging and will eventually undergo an

appendectomy. This will greatly increase the disability days for that patient. We hypothesized

that the PAT would help families better align their values and circumstances with the best

treatment for their child compared with the standardized consultation. This would then lead to

improved measures of decision making, patient-centered outcomes, and fewer disability days.

Forming the Study Cohort

Children aged 7-17 years with acute uncomplicated appendicitis and their caregivers

served as the target population for this study. This cohort represented a large group of patients

with an acute surgical disease who needed to make an urgent treatment decision between 2

treatment options with very different risk–benefit profiles (surgery vs nonoperative

management with antibiotics alone). Therefore, we screened all patients for study eligibility

based on the inclusion/exclusion criteria in Table 2. Once eligibility was confirmed, a physician

member of the research team invited eligible patients and their caregivers to participate in the

study. Informed consent and assent were obtained. Patients and their caregivers were

recruited during their initial surgical evaluation in the emergency department. Subsequently, a

trained physician member of the research team randomized the patient–caregiver dyad to

either a scripted surgical consultation (control group; standardized surgical consultation) or to a

scripted surgical consultation plus the PAT (intervention group; PAT).

20

Table 2. Inclusion/Exclusion Criteria

Inclusion Criteria Exclusion Criteria

Age: 7-17 years Ultrasound (US) or computed tomography (CT) confirmed early appendicitis:

US: hyperemia, < 1.1 cm in diameter, compressible or noncompressible, no abscess, no fecalith, no phlegmon CT: hyperemia, fat stranding, <1.1 cm in diameter, no abscess, fecalith, or phlegmon

White blood cell count < 18 C-reactive protein < 4 (if obtained) Focal abdominal pain ≤ 48 hours

• Diffuse peritonitis • Positive urine pregnancy test • Other significant comorbidities:

o Cardiovascular disease o Malignancy o Pulmonary disease o Diabetes o Severe developmental delay

After all patient and caregiver questions were answered, the patient–caregiver dyad

chose either appendectomy or nonoperative management and we informed the clinical team of

the decision. The attending surgeon spoke with the family for a full discussion of the care plan

and to obtain surgical consent if operative management was chosen. The patient was admitted

to the hospital and received care according to the standard protocols for either appendectomy

or nonoperative management of appendicitis. Any patient or caregiver who wished to speak

with a member of the surgical team prior to making a decision was provided the opportunity to

do so.

Of 242 eligible patients during the study, 42 were not enrolled. Of these 42 patients, 3

were not approached due to the study team not being contacted and 6 were not approached

because they were returning with appendicitis after a previous episode treated with antibiotics

alone. This is a prespecified clinical situation that mandates surgery based on our institutional

protocol. Of the remaining 33 eligible patients, 2 had no caregiver available to consent and 31

refused enrollment because they were too busy or did not wish to be part of research. When

considering all eligible approached patients, the overall enrollment rate for the trial was 86%

(200 of 233 patients).

21

Study Setting

We conducted this study at NCH with the resource support of the Research Institute at

Nationwide Children’s Hospital. NCH is one of the largest and busiest children’s hospitals in the

country, with a diverse patient population representing rural, urban, and minority populations

of varying educational and socioeconomic backgrounds. NCH performs more than 500

appendectomies per year. This clinical volume provided a diverse pool of patients and

caregivers to complete our trial within the proposed time period. In addition, previous studies

performed at NCH demonstrated the safety and effectiveness of nonoperative management of

uncomplicated pediatric appendicitis.33,34 Therefore, patients treated at NCH are offered the

choice between surgery and nonoperative management with antibiotics alone as part of

standard clinical practice. Finally, all baseline measurements needed to assess study eligibility

were collected as part of the routine care of children with appendicitis treated at NCH.

Interventions

Pediatric patients with appendicitis and their caregivers were randomized to either a

scripted standardized surgical consultation alone or to a scripted standardized surgical

consultation plus the PAT. Table 3 compares the content presented to each group. All

randomized patients were approached by a physician member of the research team and made

their treatment decision prior to meeting the attending surgeon.

Control Group (standardized surgical consultation)

Patient–caregiver dyads received only scripted verbal information regarding the 2

treatment options (operative or nonoperative management) using a standardized script with

patients and caregivers having the opportunity to ask questions. The patient–caregiver dyads

then chose either appendectomy or nonoperative management. Any patient or caregiver who

wished to speak with a member of the surgical team prior to making a decision was able to do

so. We developed this script with feedback from our stakeholders, to represent the best

available standard of care consultation rather than usual care.

22

Table 3. Comparison of Content Presented in the Standardized Surgical Consultation and PAT

Standardized Surgical

Consultation

PAT + Standardized Surgical

Consultation

Explained treatment options verbally x x

Explained risks and benefits verbally x x

Explained importance of patient–caregiver preferences and values when making a treatment decision

x x

Provided opportunity to ask questions x x

Provided video explanation of treatments X

Provided video explanation of risks and benefits X

Summarized side-by-side comparison of risks and benefits

X

Showed graphics that visually summarized risk and benefits (numeracy)

X

Provided patient and caregiver vignettes discussing most common concerns about treatment

X

Provided patient and caregiver vignettes discussing most common reasons patients choose each treatment

X

Provided interactive values exercise to assist patients and caregivers in prioritizing the risks and benefits for themselves

X

Provided training on the importance of and techniques for communicating with health care providers

X

Embedded activation and communication strategies X

Intervention Group (PAT)

The research physician administered the PAT plus standardized surgical consultation to

the patient–caregiver dyads and then answered their questions about the risks and benefits of

each treatment option. The PAT was designed for use by both caregivers and patients and was

23

tailored to a Flesch-Kincaid reading level between fifth and sixth grades. The patients and

caregivers were allowed to choose whether to receive the PAT together or separately, or to

have only the caregiver receive it. The PAT is a self-guided tablet-based interactive experience

(https://vimeo.com/91207174). The research physician launched the tablet experience but did

not guide the user through the experience. Once finished with the PAT, the patient–caregiver

dyad then chose either appendectomy or nonoperative management. Any patient or caregiver

who wished to speak with a member of the surgical team prior to making a decision was

provided the opportunity to do so.

Follow-up

Administering the assigned intervention and answering the questions took

approximately 20 minutes per dyad. Follow-up for both groups was identical and scheduled for

1 hour after making the treatment decision; at discharge; and at 30 days, 6 months, and 1 year

after enrollment. We wanted this study to assess the impact of the PAT on decision making and

on patient-centered and medical outcomes in both the short and long term. Therefore, we

chose to assess outcomes after the treatment decision and at the 30-day and 1-year follow-up.

Study Outcomes

Table 4 lists the outcomes, instruments, data source, and time point.

Primary Outcomes

There were 3 primary outcomes: decision self-efficacy immediately after treatment

decision, health care satisfaction at discharge, and disability days at 1 year.

24

Table 4. Study Outcomes

Measured Outcomes Measurement Instrument

Person Reporting

Time Points

Index Admission

Immediately Discharge 30 Days

6 Months

1 Year

In All Patients

Length of stay Medical chart Medical chart X

Decisional self-efficacy and confidence

Decisional Self-Efficacy Scale67

Caregiver X X

Preparedness for decision making

Preparation for Decision-Making

Scale67

Caregiver X

Caregiver activation level Parent Patient Activation

Measure®58

Caregiver X

Health care satisfaction during the initial hospitalization

PedsQL 3.0 Health care Satisfaction

Generic Module79

Caregiver X X

Quality of life (QOL) PedsQL 4.0 Generic Core Scales81

Patient and caregiver

X X

Certainty with treatment choice

Decisional Conflict Scale69

Caregiver X

Recall (knowledge) about the disease and treatment options

Study-specific knowledge survey

Caregiver X X

25

Measured Outcomes Measurement Instrument

Person Reporting

Time Points

Index Admission

Immediately Discharge 30 Days

6 Months

1 Year

Remorse or regret with treatment choice

Decision Regret Scale78

Caregiver X X

Disability days Study-specific surveys

Patient and caregiver

X X X

Readmissions Study-specific surveys/chart review

Medical chart caregiver

X X X

In Operative Patients Only

Postoperative infections Study-specific surveys/chart review

Medical chart caregiver

X

Re-operation Study-specific surveys/chart review

Medical chart caregiver

X X

Readmissions Study-specific surveys/chart review

Medical chart caregiver

X X X

In Nonoperative Patients Only

Need for appendectomy during initial admission

Study-specific surveys/chart review

Medical chart X

Recurrence Study specific surveys/chart review

Medical chart Caregiver

X X

Antibiotic complications Study specific surveys/chart review

Medical chart Caregiver

X

Abbreviation: PedsQL, Pediatric Quality of Life Inventory.

26

Decision Self-Efficacy Immediately After Treatment Decision

We assessed this in the caregiver 1 hour after the treatment decision was made using

the Decision Self-Efficacy Scale. The Decision Self-Efficacy Scale (Cronbach α ranging from 0.89

to 0.92) measures the participant’s confidence in his or her ability to make decisions. It consists

of 11 questions regarding different aspects of decision making.66-68 The scale includes 5

response categories from “not at all confident” to “very confident,” with scores ranging from 0

(extremely low self-efficacy) to 100 (extremely high self-efficacy). The scale is also correlated

with the Decisional Conflict Scale (r = 0.55) and its subscales of feeling informed and supported

and knowledge (r = 0.61).68 There is no normal range for this scale.

Health Care Satisfaction at Discharge

We assessed this by having the caregiver complete the Pediatric Quality of Life (PedsQL)

3.0 Health Care Satisfaction Generic Module-Parent Report. This survey measures the

caregiver’s satisfaction with the care his or her child received and measures 6 dimensions

(information, inclusion of family, communication, technical skills, emotional needs, and overall

satisfaction) using 24 items. Higher scores indicate higher satisfaction. The generic module was

modified from the Health Care Satisfaction Module-Parent Survey and the Pediatric

Hematology/Oncology Parent Satisfaction survey, whose Cronbach α = 0.92 and 0.96,

respectively.69,70 The modification entailed adding the phrase “at the hospital” to the final

question of the survey. There is no normal range for this scale.

Disability Days of the Child at 1 Year

We assessed this for the child at 1 year using data derived from the patient and

caregiver during interviews at discharge and at 30 days, 6 months, and 1-year follow-up, and by

review of hospital and emergency department records at each of these time points. Disability

days is a composite of inpatient hospital days, emergency department visits, primary care

physician visits, and all days with limited activities referable to appendicitis for the child

(defined as days in which the child did not participate in his or her normal routine secondary to

discomfort, symptoms, treatments, or follow-up for his or her appendicitis).

27

Secondary Outcomes

We categorized secondary outcomes as either decision making and patient-centered

outcomes or medical outcomes.

Decision Making and Patient-Centered Outcomes (Table 4)

We utilized the different scales detailed below to assess various components of decision

making. The decision self-efficacy, preparedness for decision making, decisional conflict, and

decision regret scales all reflect different measures of the caregivers’ perspective of decision

making, including how well informed they are about treatment options and whether the

presented information facilitates a confident decision. The Parent Patient Activation Measure

(Parent-PAM) measures if the PAT increases caregiver activation level.

Decision Self-Efficacy at Discharge and 30 Days

We assessed this by having the caregiver complete the Decision Self-Efficacy Scale

(described previously).

Preparation for Decision Making Immediately After Treatment Decision

We assessed this in the caregiver using the Preparation for Decision Making Scale. The

Preparation for Decision Making Scale (Cronbach α ranging from 0.92 to 0.96) is a 10-item scale

used to assess the patient’s perception of the usefulness of a decision support intervention in

preparing to make a health decision after a patient has discussed his or her treatment

options.67,71-73 There are 5 response categories ranging from 1 (not at all) to 5 (a great deal).

The scale is scored by summing the 10 items and dividing by 10; higher scores indicate higher

perceived levels of preparation for decision making.67 The test has been used to compare the

effects of different decision support interventions in previous studies.71,72 There is no normal

range for this scale.

28

Decisional Conflict Immediately After Treatment Decision

We assessed this in the caregiver using the Decisional Conflict Scale. The Decisional

Conflict Scale (Cronbach α > 0.8) measures (1) uncertainty in choosing options and (2)

modifiable factors contributing to uncertainty.74 The scale comprises 10 items with 4 subscales

(informed, values clarity, support, and uncertainty) and 3 response categories: yes, no, and

unsure.74 Scores range from 0 (no decisional conflict) to 100 (extremely high decisional

conflict).75 This scale has been used in various studies and is measured immediately after a

decision has been made.76-79 There is no normal range for this scale.

Decision Regret Scale at Discharge and at 30-day Follow-up

We assessed this in the caregiver using the Decision Regret Scale. The Decision Regret

Scale (Cronbach α ranging from 0.81 to 0.92) is a 5-item scale that measures “distress or

remorse after a (health care) decision” with 5 response categories from “strongly agree” to

“strongly disagree.”80,81 The scale is given to participants at some point after they have made a

decision and are able to reflect on its effects. The scale has been reliable across various patient

populations and health care decisions.81-83 The scale has correlated well with satisfaction with

the decision (r = –0.40 to –0.60), decisional conflict (r = 0.31 to 0.52), and overall rated quality

of life (r = –0.25 to –0.27).81 There is no normal range for this scale.

Parent-PAM Immediately After Treatment Decision

We assessed this in the caregiver using a version of the healthy Child–Parent-PAM,

adapted with permission. We adapted it to include the specific treatment options in question 4

and the specific disease “appendicitis” in question 9. The Parent-PAM is a 13-item scale and

was derived from the short-form version of the Patient Activation Measure.50 The Parent-PAM

assesses the parent’s activation in the management of his or her child’s health. This is scored on

a Likert-type scale with 4 response categories ranging from 1 (disagree strongly) to 4 (agree

strongly) and a Cronbach α of 0.85.50 The scores are summed from 0 to 100, with higher scores

indicating higher activation. The Parent-PAM has 4 levels of activation: Level 1—does not

believe the caregiver role is important (score ≤ 47.0); Level 2—does not have the confidence or

29

skill to take action (score ≥ 47.1 and ≤ 55.1); Level 3—may begin to take action (score ≥ 55.2

and ≤ 67.0); and Level 4—takes action but may have difficulty maintaining behaviors over time

(score ≥ 67.1).4 There is no normal range for this scale.

Knowledge/Recall Immediately After Treatment Decision and at 30-day Follow-up

We assessed this in the caregiver using an internally created 4-question survey that

assessed the caregiver’s knowledge about appendicitis. We created this survey using

stakeholder input.

HRQOL of the Patient at Discharge and at 30-day Follow-up

We assessed this for the child using both self-reports and caregiver proxy reports. We

used validated questionnaires from the Pediatric Quality of Life Inventory (PedsQL) scales.

We chose the PedsQL because it is validated, is available in multiple languages, and has

been used to study pediatric patients with appendicitis.84,85 We specifically used the PedsQL 4.0

Generic Core Scales—Child and Parent Report. The generic core scales module measures 4

dimensions (physical, emotional, social, and school functioning) using 23 items and is designed

to measure the HRQOL of healthy or ill (acute or chronic) children through self-reports and

parent proxy reports. The generic core scales are valid and reliable with internal consistency

values (Cronbach α) of 0.88 for the child report and 0.90 for the parent proxy report.86-88 These

scales also are interpreted similarly across pediatric race/ethnic groups, socioeconomic groups,

and in various modes of administration.89-91 There is no normal range for these scales.

Health Care Satisfaction at 30-day Follow-up

We assessed this in the caregiver using the PedsQL 3.0 Health Care Satisfaction Generic

Module-Parent Report described above.

30

Medical Outcomes (Table 4)

Medical outcomes include disability days, length of hospital stay, medical/surgical

complications, recurrence, and hospital readmissions. We assessed these using data derived

from the patient and caregiver during interviews at discharge and at 30-day, 6-month, and 1-

year follow-up and by reviewing hospital and emergency department records at each of these

time points. We assessed disability days 30 days, 6 months, and 1 year after the diagnosis of

appendicitis. Additional secondary medical outcomes included standard surgical outcomes that

are measured and reported in clinical trials, including hospital readmissions, emergency

department visits, and length of stay.

Data Collection and Sources

Figure 1 depicts the overall study design and data collection. We obtained both baseline

data and outcome data from medical record review and from patient–caregiver interview. To

minimize participant dropout and loss to follow-up, we offered options of in-person, phone, or

email follow-up. Research staff contacted participants using multiple methods and also sent

reminder emails to all participants between follow-up time points. In addition, patient–

caregiver dyads received incentives for continued participation in the study through 1-year

follow-up ($25 Visa gift cards at 30-day, 6-month, and 1-year follow-up). We maintained

current contact information for all participants by contacting them quarterly through phone

and email queries. No patients told us they were withdrawing from the study. Loss to follow-up

was due to an inability to contact patients–caregivers based on contact information provided to

study staff and listed in the medical record. Therefore, we do not have specific reasons for all

participants’ dropout.

31

Figure 1. Study design

Abbreviations: OHRI, Ottawa Hospital Research Institute; PedsQL, Pediatric Quality of Life Inventory.

32

Analytical and Statistical Approach

Sample Size and Power

We enrolled 200 patients in order to have sufficient power to analyze the 3 primary

outcomes. The primary patient-centered outcomes are decision self-efficacy immediately after

the treatment choice and health care satisfaction at discharge. We expected the PAT to

increase the total Decision Self-Efficacy Scale score by 0.4 standard deviations and increase

total PedsQL Health Care Satisfaction Generic Module scores at discharge by 0.5 standard

deviations.92,93 For the total Decision Self-Efficacy Scale, a higher score implies a higher self-

efficacy. The maximum score of 100 reflects very high confidence, a score of 50 reflects

moderate confidence, and the minimum score of 0 reflects no confidence at all. For the total

PedsQL Health Care Satisfaction score, a higher score indicates higher satisfaction. The

maximum score of 100 reflects very high satisfaction, a score of 50 reflects moderate

satisfaction, and the minimum score of 0 reflects minimal satisfaction. Assuming that we would

show superiority for at least 1 of these patient-centered outcomes, we also considered it critical

to assess the effects of the PAT on disability days compared with standardized surgical

consultation.

Based on our previous research and published data, we expected the number of

disability days in the standardized surgical consultation group to be 14 days during the first year

of follow-up with a standard deviation of 5 days.94 Based on this, we set our noninferiority

margin to 14%, or 2 days more. For the primary patient-centered outcomes, we planned to use

Holm’s procedure to test the 2 primary outcomes. Noninferiority of disability days would be

meaningful only if we achieved significance on these primary outcomes. Based on this, we

planned to use a gatekeeping fallback procedure, whereby we could test for noninferiority at

the 1-sided 0.025 level when we achieve significance for 1 or both of the patient-centered

primary outcomes.95 If we failed to reach significance on these primary outcomes, then we

planned to use a much more stringent alpha of 0.005 (fallback) to test for noninferiority of the

33

PAT for disability days. We proposed the more stringent alpha level for this test because we

specified, a priori, that differences between groups in disability days would be relevant only if

we found a significant difference between groups in either decision self-efficacy or health care

satisfaction. Per this supposition, a gatekeeping fallback procedure using a more stringent alpha

was necessary for the test of difference in disability days between groups in order to preserve

the overall type 1 error for the testing approach at 0.05. Based on this strategy, 100 patients in

each group were needed to provide power of 92% to reach significance on either primary

patient-centered outcome, 74% power to reach significance on both, and 80% power to claim

significant noninferiority for disability days.

We planned for an interim analysis after 12 months of enrollment (50% of sample size

enrolled) to assess for (1) effects of the PAT on the rates of patient choice of the 2 treatment

options (selection bias), (2) negative effects of the PAT on medical outcomes and treatment-

associated disability, and (3) negative effects of the PAT on patient-reported outcomes.

Planned Analyses: Decision Making and Patient-Centered Outcomes

The primary and secondary decision making and patient-centered outcomes are

detailed in Table 4 and Figure 1. Several outcomes measured at 2 or 3 time points (eg, decision

self-efficacy within 1 hour after the intervention, prior to discharge, and 30 days postdischarge)

would be analyzed using linear mixed models. Patient-centered outcomes to be analyzed using

mixed models across all time points at which the measure was determined were total decision

self-efficacy scores, total PedsQL Health Care Satisfaction Generic Module scores, and total

PedsQL HRQOL General Core Scale scores in patients and caregivers. We included time as a

categorical variable in each of these models, as well as its interaction with the PAT intervention.

The interaction terms would provide estimates of the differences in these patient-centered

outcomes between PAT and standardized surgical consultation groups at each time point in the

study. We planned to use an unstructured error covariance approach to account for the within-

subject correlation in outcomes measured over time.

34

Planned Analyses: Medical Outcomes

The primary and secondary medical outcomes are detailed in Table 4 and Figure 1.

Based on a meta-analysis of trials of nonoperative management of acute uncomplicated

appendicitis and our local institutional data, we expected those who received the standardized

surgical consultation to have an average of 14 (standard deviation of 5) disability days at 1 year

(and at 30 days and 6 months).22 We also expected a 15% readmission rate in the group that

received the standardized surgical consultation. For disability days in a year, we planned to test

for noninferiority with a margin of less than 2 days. We chose the 2-day margin because it was

the maximum increase in disability days that was considered reasonable by our clinician and

patient/caregiver stakeholders. Any larger increase in disability days would likely prohibit the

PAT from being incorporated into clinical practice. It was also possible that the PAT could

reduce disability days. The noninferiority test could be followed by a superiority test for

disability days, with no effect on type 1 error.96 In other words, if the PAT happened to

decrease days significantly as compared with standardized surgical consultation, a strong

conclusion about superiority would have been allowed despite designing the study as a

noninferiority trial.96-99 The PAT and standardized surgical consultation groups would also be

compared on their average readmission rates. We planned to assume a negative binomial

distribution using the GENMOD procedure in SAS 9.3 to model the count of days.

Planned Exploratory Subgroup Analyses

We performed exploratory subgroup analyses to examine outcomes relevant only to the

operative or nonoperative treatment groups. This would have been especially useful if we

found a difference in disability days between the randomized groups because the difference

could have been related to differences in the percentage of patients within the PAT group

compared with the standardized surgical consultation group that choose operative vs

nonoperative management. Within the subgroup of patients that chose operative therapy, we

compared postoperative infections, re-operation rate at 1 year, and readmission rate at 1 year

to identify if receiving the PAT altered these medical outcomes. Within the subgroup of patients

who chose nonoperative therapy, we examined the need for appendectomy during initial

35

admission, recurrence of appendicitis at 1 year, and antibiotic adverse effects at 30 days to

identify if receiving the PAT altered these outcomes. We compared proportions for antibiotic

complications, recurrence, postoperative infection, and complications. We also compared

average readmission rates using a Fisher exact test.

Planned Analyses for Missing Data

The potential for missing data is a concern for any longitudinal study. We expected a

small percentage of patients to miss follow-up visits or phone calls, but we aimed to keep this

percentage to less than 10% at each follow-up time point. We attempted to reduce the amount

of missing data through phone call reminders prior to the 30-day follow-up appointments, clear

instructions at discharge about future study follow-up visits and phone calls, and compensation

for completing follow-up assessments. To handle missing data, we planned to use multiple

imputation by chained equations (MICE) in analyses of follow-up data. This statistical

methodology is a multivariate imputation of missing values under a missing at random (MAR)

assumption. The MAR assumption is a much less strict assumption than the missing completely

at random (MCAR) assumption, which would be required if we were to analyze only those

patients with complete data. The MAR assumption might be too strong also. Therefore, we

used recommended sensitivity analysis to determine the impact on the estimated treatment

effect by differing assumptions about an informative missing data mechanism.100

Analyses Performed

We included all randomized participants in the analyses. Initial analyses included

generating summary statistics using frequency (percentage) for categorical variables and

median (interquartile range [IQR]) for continuous variables. We tested initial comparisons of

dichotomous outcomes using Fisher exact tests. We tested continuous outcomes using

Wilcoxon rank sum tests. Planned heterogeneity of treatment effect (HTE) analyses described in

our application included planned subgroup analyses by treatment choice. We compared rates

of postoperative infections, reoperation, and readmission at 1 year among operative patients

who did and did not receive the PAT. We also compared, among patients who chose antibiotics,

36

the rates of appendectomy during the initial admission, recurrence of appendicitis at 1 year,

and antibiotic adverse effects at 30 days between those who did and did not receive the PAT.

These analyses and hypotheses for these HTE analyses were prespecified. We compared

proportions for antibiotic complications, recurrence, readmission, postoperative infection, and

complications. In addition, for the 3 primary outcomes, we performed subgroup comparisons

based on the randomized intervention and treatment choice using Kruskal-Wallis tests for

continuous variables and Fisher exact tests for categorical variables. We did not perform the

additional planned analyses detailed above, as there were no significant differences in favor of

the PAT for any of the primary or secondary outcomes.

For baseline data and each completed follow-up, there were minimal missing data.

There were several patients with loss to follow-up, with follow-up rates for the patients who

chose surgery in both randomized groups below our target of 90%. We do not believe there

was a preferential loss to follow-up based on the randomized group, as both the PAT and

standardized consultation groups had similar proportions of lost to follow-up patients who

chose surgery. It is expected that patients who chose surgery for appendicitis would be more

likely to be lost to follow-up. These patients have resolution of their disease once recovered

from surgery and would be more likely to not want to continue participating in ongoing follow-

up. Therefore, we did not use the prespecified strategies for handling missing data outlined in

the original application.

Conduct of the Study

The IRB at NCH approved this study on July 12, 2013, and remains active until the

dissemination of the study results. During the design phase of the PAT, the stakeholders

strongly suggested that we revise our enrollment process to have a physician perform the

enrollment and administer the assigned intervention rather than a research nurse (as described

in the original proposal). We made this change to our IRB application prior to beginning the

trial, and a physician member of the research team (research fellow) performed all enrollments

and administered the randomized intervention throughout the trial. The use of the research

37

fellows to administer both interventions may have led to contamination of the interventions. It

is possible that the research fellows unknowingly included aspects of the PAT in their verbal

communication with patients and caregivers randomized to the control group.

Throughout the remainder of the study, the study protocol remained constant and did

not change from the originally proposed protocol. We made no changes to the study

populations, data sources, or methodological comparators.

RESULTS

Enrollment and Follow-up

This RCT of the effects of a PAT in pediatric appendicitis enrolled patients with

uncomplicated appendicitis from March 2014 through April 2016. Figure 2 details the study

flow. We screened all 1114 patients with appendicitis treated at NCH during the study period,

with 914 patients being excluded. Of the excluded patients, 872 did not meet inclusion criteria

and 42 were excluded for other reasons (Table 5). The overall enrollment rate of eligible

approached patients was 86% (200 of 233 patients).

38

Table 5. Reasons for Exclusion of 914 From the Studya

Excluded Because Did Not Meet Inclusion/Exclusion Criteria (n = 872)b

Number Screened

% Screened

Age < 7 years 133 15

Age > 17 years 16 2

Symptoms > 48 hours 212 23

White blood cell (WBC) ≤ 5 54 6

WBC ≥ 18 289 32

Appendix > 1.1 cm in diameter 371 41

Appendicolith 446 49

Evidence of rupture 272 30

History of chronic abdominal pain 34 4

Not initially diagnosed with appendicitis 21 2

Excluded for Other Reasons (n = 42) Number Excluded

% Excluded

Too busy/did not wish to be bothered 31 74

Returning with appendicitis after previous nonoperative management

6 14

Emergency department staff failed to contact the study team

3 7

Patient with no available guardian to consent 2 5 a Numbers do not sum to 914 because many patients were excluded for multiple reasons. b Numbers do not sum to 872 because many patients were excluded for multiple reasons.

39

Figure 2. CONSORT flow diagram

Assessed for eligibility (n=1114)

Excluded (n=914) ♦ Not meeting inclusion criteria (n=872) ♦ Recurrent appendicitis (n=6) ♦ Not approached (n=3) ♦ Declined to participate (n=31) ♦ No guardian available to consent (n=2)

♦ Lost to follow-up at 30day follow-up (n=15) ♦ Lost to follow-up at 6 month follow-up (n=15) ♦ Lost to follow-up at 1 year follow-up (n=13)

Allocated to Patient Activation Tool (n=98) ♦ Received allocated intervention (n=98) ♦ Did not receive allocated intervention (give

reasons) (n= 0)

Allocated to Standard Consultation (n=102) ♦ Received allocated intervention (n=102) ♦ Did not receive allocated intervention (give

reasons) (n=0)

Allocation

Analysis

Follow-Up

Randomization (n=200)

Enrollment

Management Choice

Choice: Operative (n=67) ♦ Received choice: (n=67) ♦ Did not receive choice

(n=0)

Choice: Non-Operative (n=31) ♦ Received choice: (n=31) ♦ Did not receive choice (n=0)

Choice: Operative (n=60) ♦ Received choice: (n=60) ♦ Did not receive choice

(n=0)

Choice: Non-Operative (n=42) ♦ Received choice: (n=42) ♦ Did not receive choice

(n=0)

Analysed at 30day follow-up (n=52) Analysed at 6 month follow-up (n=52) Analysed at 1 year follow-up (n=54) ♦ Excluded from analysis (n=0)

♦ Lost to follow-up at 30day follow-up (n=4) ♦ Lost to follow-up at 6 month follow-up (n=4) ♦ Lost to follow-up at 1 year follow-up (n=2)

♦ Lost to follow-up at 30day follow-up (n=13) ♦ Lost to follow-up at 6 month follow-up (n=9) ♦ Lost to follow-up at 1 year follow-up (n=8)

♦ Lost to follow-up at 30day follow-up (n=3) ♦ Lost to follow-up at 6 month follow-up (n=3) ♦ Lost to follow-up at 1 year follow-up (n=3)

Analysed at 30day follow-up (n=27) Analysed at 6 month follow-up (n=27) Analysed at 1 year follow-up (n=29) ♦ Excluded from analysis (n=0)

Analysed at 30day follow-up (n=47) Analysed at 6 month follow-up (n=51) Analysed at 1 year follow-up (n=52) ♦ Excluded from analysis (n=0)

Analysed at 30day follow-up (n=39) Analysed at 6 month follow-up (n=39) Analysed at 1 year follow-up (n=39) ♦ Excluded from analysis (n=0)

40

Of the 200 patients enrolled, 98 were randomized to the PAT (intervention group) and

102 to the standardized consultation group. Among the PAT group patient–caregiver dyads,

76% received the PAT together, 2% received the PAT separately, and 22% had the caregiver

only receive the PAT. Within the PAT group, 67 chose surgery and 31 chose nonoperative

management. Within the standardized consultation group, 60 chose surgery and 42 chose

nonoperative management. We found no difference in the proportion of patients who chose

nonoperative management or surgery between the PAT (32% antibiotics and 68% surgery) and

standardized consultation groups (41% antibiotics and 59% surgery) (p = 0.19). Follow-up rates

at 30 days, 6 months, and 1 year by randomized group were similar between the PAT (80.6% at

30 days, 80.6% at 6 months, and 84.7% at 1 year) and standardized consultation group (84.3%

at 30 days, 88.2% at 6 months, and 89.2% at 1 year). Follow-up rates by randomized group and

treatment choice are shown in Table 6. We excluded no patients from any of the analyses.

41

Table 6. Follow-up Rates by Randomized Group and Treatment Choice

Randomized group 30 Day 6 Month 1 Year

Completed % Completed % Completed %

PAT (n = 98) 79 81 79 81 83 85

Standardized consultation (102)

86 84 90 88 91 89

Treatment choice

Surgery (n = 127) 99 78 103 81 106 84

Nonoperative management (n = 73)

66 90 66 90 68 93

Randomized group and treatment choice

PAT + surgery (n = 67) 52 78 52 78% 54 81

PAT + nonoperative management (n = 31)

27 87 27 87 29 94

Standardized consultation + surgery (n = 60)

47 78 51 85 52 87

Standardized consultation + nonoperative management (n = 42)

39 93 39 93 39 93

The baseline demographics and clinical characteristics for each group are shown in Table

7. The groups appear to be similar in all measured variables.

42

Table 7. Baseline Demographics and Clinical Characteristics for the 2 Groups

Characteristic Total (n = 200)

Standardized Consultation

(n = 102)

PAT (n = 98)

Hospital length of stay, hours Median (interquartile range [IQR]) 23 (16-32) 24 (17-34) 21 (16-30)

Age, years [Median (IQR)] 12 (9-15) 13 (9-15) 12 (10-14)

Gender [n (%)] 120 (60.0) 61 (59.8) 59 (60.2)

Male

Race [n (%)] 173 (86.5) 85 (83.3) 88 (89.8)

White

Primary language other than English [n (%)] 21 (10.5) 10 (9.8) 11 (11.2)

Insurance status [n (%)]

Private 126 (63.0) 67 (65.7) 59 (60.2)

Medicaid 65 (32.5) 28 (27.5) 37 (37.8)

No insurance 9 (4.5) 7 (6.9) 2 (2.0)

Presenting symptoms [n (%)]

Fever 41 (20.5) 21 (20.6) 20 (20.4)

Vomiting 86 (43.0) 45 (44.1) 41 (41.8)

Diarrhea 19 (9.5) 9 (8.8) 10 (10.2)

Diagnostic imaging performed [n (%)]

Ultrasound 158 (79.0) 79 (77.5) 79 (80.6)

CT scan 56 (28.0) 27 (26.5) 29 (29.6)

White blood cell (WBC), 1000 cells/mcL [Median (IQR)] 11.8 (9.6-14.6) 11.8 (9.0-14.5) 11.7 (9.8-15.0)

Duration of abdominal pain, hours [Median (IQR)] 18 (12-24) 18 (12-24) 17 (12-24)

43

Interim Analyses

Interim results after 12 months of enrollment indicated that (1) there did not appear to

be a treatment selection bias incurred by the use of the PAT, (2) there did not appear to be any

negative effects of the PAT on medical outcomes and treatment-associated disability, and (3)

there did not appear to be any negative effects of the PAT on patient-reported outcomes. For

the primary patient-centered outcomes—decision self-efficacy scores and the PedsQL Health

Care Satisfaction Generic Module scores—the interim analysis did suggest a lack of benefit of

the PAT on either outcome. However, for the primary medical outcome of disability days, the

interim analysis suggested that the PAT could lead to fewer disability days. We reviewed these

interim results and allowed the trial to enroll to completion, as additional enrollment and

completion of 1-year follow-up was needed to (1) achieve sufficient power to perform our

planned sequential analyses for noninferiority and potential superiority of the PAT over

standard surgical consultation on disability days; (2) allow for complete analysis of the 2

patient-centered primary outcomes, including use of mixed models to assess total decision self-

efficacy scores and total PedsQL Health Care Satisfaction scores, and to assess for interactions

between the outcomes over time; (3) allow for testing of an interaction between the PAT and

standard surgical consultation and the 2 treatment options pertaining to medical and patient-

centered outcomes; and (4) perform the exploratory subgroup analyses specified in our

application.

Primary Outcomes

Decisional self-efficacy within 1 hour was significantly higher in the standardized

consultation group compared with the PAT group (median [IQR]: 100 [97.7-100] vs 100 [95.5-

100]; p = 0.03) with 70% of the standardized consultation group scoring 100 compared with

55% of the PAT group scoring 100 (Table 8). Figure 3 displays the distribution of scores on the

Decisional Self-Efficacy Scale at 1 hour. The very high median scores in each group potentially

reflect a ceiling effect of the survey instrument. In addition, the equivalent median scores of

100 for both groups indicates that each group had “extremely high self-efficacy” based on this

scale; therefore, the detected statistical difference is not likely clinically meaningful. We

44

performed additional post hoc exploratory analyses to determine if the PAT led to a lower

proportion of patients with low decision self-efficacy compared with standardized consultation.

We examined the proportion of patients scoring greater than potentially clinically relevant

cutoffs based on the scale. When using cutoff scores of 50% (reflecting moderate self-efficacy

on the scale) and 75% (reflecting high self-efficacy on the scale), we found no significant

differences between the 2 groups (Table 8).

Health care satisfaction at discharge assessed using the PedsQL 3.0 Health Care

Satisfaction Generic Module-Parent Report was similarly high in both groups, with no significant

difference between the standardized consultation group compared with the PAT group (99

[94.7-100] vs 98 [91.7-100]; p = 0.27). (Table 8) The very high median scores in each group

potentially reflect a ceiling effect of the survey instrument. Disability days at 1 year were also

similar between the standardized consultation and PAT groups (6 [2-11] vs 5 [2-15]; p = 0.67)

(Table 8).

45

Table 8. Primary Outcomes

Outcome Total Standardized Consultation PAT P Value

n % Median IQR n % Median IQR n % Median IQR

Decisional Self-Efficacy Scale

Immediately after decision

— — 100 95.5-100 — — 100 97.7-100 — — 100 95.5-100 0.03

> 50% 195 99.5 — — 101 100 — — 94 98.9 — — 0.48

> 75% 190 96.9 — — 98 97 — — 92 96.8 — — 1

Health care satisfaction during initial hospitalization (PedsQL 3.0, health care satisfaction)

At discharge — — 98.9 94-100 — — 99 94.7-100 — — 98.3 91.7-100 0.27

Disability days

1 year follow-up — — 5.5 2-12 — — 6 2-12 — — 5 2-15 0.67

Abbreviations: IQR, interquartile range; PAT, patient activation tool; PedsQL, Pediatric Quality of Life Inventory.

46

Figure 3. Distribution of scores on the Decisional Self-Efficacy Scale at 1 hour

Abbreviation: PAT, patient activation tool.

Decision Making and Patient-Centered Outcomes

The results of all other decision making and patient-centered outcomes assessed

throughout the trial are shown in Table 9. Decisional self-efficacy scores were similar between

the groups at discharge and 30 days. Scores on the Preparation for Decision Making Scales were

significantly lower in the PAT group compared with the standardized consultation group,

immediately after the decision (95 [80-100] vs 100 [87.5-100]; p = 0.03). The very high median

scores in each group reflect being “a great deal” prepared for decision making; therefore, the

detected statistical difference is not likely clinically meaningful. We found no significant

differences between the 2 groups on measures of parent patient activation or decisional

conflict immediately after the decision. We found no significant differences in HRQOL, health

care satisfaction, decisional regret, or guardian knowledge/recall at any time point.

47

Table 9. Decision Making and Patient-Reported Outcomes

Patient-reported Outcome Time Point Standardized Consultation

PAT P Value

N Total

Median (IQR) Median (IQR)

Decisional Self-Efficacy Scale At discharge 100 (97.7-100) 100 (97.7-100) 0.41 193

30 days 100 (95.5-100) 100 (90.9-100) 0.38 150

Preparation for Decision Making Scale

Immediately after decision

100 (87.5-100) 95 (80-100) 0.03 197

Parent Patient Activation Measure (caregiver activation level)

Immediately after decision

100 (77.7-100) 91 (75-100) 0.17 196

Health care satisfaction during initial hospitalization (PedsQL 3.0 Health Care Satisfaction)

30 days 99.0 (92.8-100) 96.9 (84.5-100) 0.18 150

Quality of life - Patient (PedsQL 4.0 Generic Core Module)

At discharge 89.1 (81.5-94.6) 89.1 (77.2-94.6) 0.39 193

30 days 88.0 (81.0-96.7) 88.0 (79.3-96.7) 0.99 147

1 year 93.5 (85.3-98.9) 92.9 (83.2-98.9) 0.49 164

Quality of Life - Parent (PedsQL 4.0 Generic Core Module)

At discharge 90.2 (82.6-94.6) 89.7 (78.3-96.7) 0.71 192

30 days 90.2 (80.4-97.8) 87.0 (77.2-94.6) 0.21 149

1 year 94.6 (87.0-100) 92.4 (85.9-97.8) 0.25 169

Decisional Conflict Scale

Immediately after decision

0 (0-0) 0 (0-0) 0.42 197

Decision Regret Scale

At discharge 40 (40-45) 40 (40-45) 0.32 192

30 days 40 (40-40) 40 (40-40) 0.58 163

Guardian knowledge recall assessment, % correct

Immediately after decision

100 (100-100) 100 (100-100) 0.18 200

30 days 100 (100-100) 100 (100-100) 0.11 163

48

Medical Outcomes

The results of all other medical outcomes assessed throughout the trial are shown in

Table 10. At 30 days postdischarge, we found no significant differences in any of the measured

medical outcomes between standardized consultation and PAT groups, including rates of early

treatment failure, complicated appendicitis, hospital readmission, emergency

department/urgent care visits, postoperative complications, or disability days. Patients in the

PAT group had a higher rate of fever postdischarge; the rates of all other postdischarge

symptoms were similar between the 2 groups. At 6 months postdischarge, we found no

significant differences in any of the measured medical outcomes between standardized

consultation and PAT groups including hospital readmission, emergency department/urgent

care visits, disability days, treatment failure/recurrence, or complicated appendicitis. At 1 year

postdischarge, we found no significant differences in any of the measured medical outcomes,

including hospital readmission, emergency department/urgent care visits, disability days,

treatment failure/recurrence, or complicated appendicitis, between standardized consultation

and PAT groups. The nonmedical components of disability days (missed school days and normal

activity days missed) are also detailed in Table 10. Among all patients who underwent surgery

(primary choice plus failure of nonoperative management), we found no differences between

the standardized consultation group and the PAT group in the proportion of patients with

complicated appendicitis at any time. Among patients who chose nonoperative management,

we found no difference in the treatment failure rate of nonoperative management at any time

between standardized consultation and PAT groups.

49

Table 10. Medical Outcomes at 30 Days, 6 Months, and 1 Year

Standardized Consultation

PAT P Value

n % n %

Medical Outcomes 30 Days Postdischarge 86 79

Early treatment failure, within 30 days (nonoperative only)a

42 30

Any 7 16.7 2 6.7 0.29

Complicated 2 4.8 1 3.3 1.00

Complicated appendicitis (all surgery or treatment failure/recurrence)

68

71

7 10.3 9 12.7 0.79

Hospital encounter 86 79

Hospital readmission b 8 9.2 7 8.9 1.00

Emergency department/urgent care visit 4 4.7 5 6.3 0.74

Symptoms postdischarge 86 79

Fever 0 0 5 6.3 0.02

Abdominal pain 21 24.4 20 25.3 1.00

Nausea 5 5.8 4 5.1 1.00

Vomiting 1 1.2 0 0 1.00

Diarrhea 11 12.8 11 13.9 1.00

Constipation 4 4.7 5 6.3 0.74

Rashes 3 3.5 3 3.8 1.00

Postoperative complications (surgery or early treatment failure)

54

53

Wound infection 1 1.9 0 0 1.00

Intra-abdominal infections 0 0 0 0 n/a

Urinary tract infection 0 0 0 0 n/a

50

Standardized Consultation

PAT P Value

n % n %

Pneumonia 0 0 0 0 n/a

Medical Outcomes 6 Months Postdischarge 90

79

Treatment failure/recurrence within 6 monthsa

40 30

Any 13 32.5 8 26.7 0.79

Complicated 2 5 1 3.3 1.00

Hospital encounter 90 79

Hospital readmissionb 15 16.5 12 15.2 0.84

Emergency department/urgent care visit 5 5.6 5 6.3 1.00

Medical Outcomes 1 Year Postdischarge 91

83

Treatment failure/recurrence within 1 yeara 38 30

Any 13 34.2 11 36.7 1.00

Complicated 2 5.3 1 3.3 1.00

Hospital encounter 91 83

Hospital readmissionb 15 16.3 15 18.1 0.84

Emergency department/urgent care visit 6 6.6 6 7.2 1.00

Disability Days 30 Days Postdischarge Median IQR Median IQR P Value

Disability days 5.5 2-11 5 2-10 0.68

School days missed 2 1-4 2 1-4 0.60

Normal activity days missed 4.5 1-10 3.5 1-9 0.79

Guardian days missed from normal activities 2 1-4 2 1-4 0.93

Medical Outcomes 6 Months Postdischarge

51

Standardized Consultation

PAT P Value

n % n %

Disability days 5.5 2-11 5 1-12 0.90

School days missed 2 1-4 2.5 1-5 0.51

Normal activity days missed 5 1-10 4.5 1-12 0.96

Guardian days missed from normal activities 2 1-4 2 1-4 0.74

Medical Outcomes 1 Year Postdischarge

Disability days 6 2-11 5 1-15 0.67

School days missed 2 1-4 3 1-5 0.36

Normal activity days missed 5 1-10 5.5 1-14 0.76

Guardian days missed from normal activities 2 1-4 2 1-4 0.43

Note: Dichotomous outcomes tested by Fisher exact test. Continuous outcomes tested by Wilcoxon rank sum test. a Failure/recurrence sample sizes include all patients having treatment failure or recurrence on or before the follow-up time point and/or having available follow-up data at or after the follow-up time point. b Readmission sample sizes include all patients with data reported at the follow-up time point or a treatment failure or recurrence on or before the follow-up time point.

Effect of Treatment Choice on Primary Outcomes Based on Randomized Group (HTE)

As part of the planned assessment for HTE, we examined the primary outcomes based

on randomized group and treatment choice (Table 11). We found a significant difference in

decisional self-efficacy immediately after the decision across the 4 groups, with patients in the

PAT antibiotics group having the lowest median score (Standardized Consult—Surgery 100,

Standardized Consult—Antibiotics 100, PAT—Surgery 100, PAT—Antibiotics 97.7; p = 0.02). The

very high median scores in each of the 4 groups demonstrate that each group had “extremely

52

high self-efficacy.” This raises the potential for a ceiling effect, with the detected statistical

difference not likely being clinically meaningful. Additional analyses examining the proportion

of patients scoring greater than 50% (reflecting moderate self-efficacy) and 75% (reflecting high

self-efficacy) demonstrated no significant differences between the groups (Table 11). We found

no significant differences based on the randomized group and treatment choice in disability

days at 1 year or health care satisfaction at discharge.

53

Table 11. Primary Outcomes by Randomized Group and Treatment Choice

Outcome Standardized Consultation PAT P Value

Surgery Antibiotics Surgery Antibiotics

n % Median IQR n % Median IQR n % Median IQR n % Median IQR

Decisional self-efficacy immediately after treatment decision

Immediately after treatment

59 — 100 100-100 42 — 100 95.5-100 66 100 95.5-100 29 — 97.7 93.2-100 0.02

> 50% 59 100 — — 42 100 — — 65 98.5 — — 29 100 — — 1.00

> 75% 58 98.3 — — 40 95.2 — — 64 97.0 — — 28 96.6 — — 0.85

Health care satisfaction

At discharge 45 — 99 89.6-100 33 — 99 95.9-100 46 — 97.4 87.5-100 26 — 93.9 76.1-100 0.35

Disability days

1 year 53 — 8 2-15 38 — 4 2-7 53 — 8 3-15 30 — 3 1-16 0.10

Abbreviations: IQR, interquartile range; PAT, patient activation tool.

54

DISCUSSION

Decisional Context

The goal of this study was to determine if a PAT can activate pediatric patient–caregiver

dyads to make more confident informed decisions about a treatment course for appendicitis.

We attempted to determine whether a PAT can be successfully utilized in an emergency care

situation to improve decision making and patient-centered outcomes (activation, knowledge,

preparedness of decision making, decision self-efficacy and regret, decisional conflict,

satisfaction, and HRQOL) without compromising medical outcomes (disability days, readmission

rates, and length of hospital stay). Overall, we found that patients in both groups had high

scores on tests that measure their ability and comfort level to make a treatment decision for

their child. We did not detect any difference between the PAT and standardized consultation

alone in measures of decision making, patient-centered outcomes, or medical outcomes. The

PAT did not lead to increased harm compared with standardized consultation alone.

For patients and families making a medical treatment decision in an emergency care

setting, our results demonstrated no difference in measures of decision making or patient

activation between a technology-based tool and a standardized consultation that emphasized

patient choice. It is possible that an intervention to improve decision making in this setting may

not be needed.

Study Results in Context

Prior to this study, there was little evidence examining patient activation on decision

making and patient-centered outcomes in the emergency care setting, or in pediatric surgery.

This study specifically investigated the effects of patient activation on decision making and

patient-centered outcomes in an acute surgical disease in children. Previous adult studies have

demonstrated benefits of patient communication interventions on outcomes, particularly in the

outpatient setting. In addition, decision aids have been shown to improve measures of decision

making in adult patients. To our knowledge, this was the first study to combine a patient

55

activation strategy with a decision aid for patient–caregiver dyads in the emergency surgical

setting.

This study did not find a difference in measures of decision making, patient-centered

outcomes, or medical outcomes between a PAT—designed to activate caregivers in the

emergency setting through knowledge and engagement—and a standardized consultation that

emphasized patient choice. This study demonstrated extremely high scores on measures of

decision making in both groups.

Implementation of Study Results

Patients, families, physicians, and hospital administrators will be the key end users of

our results. Incorporating patients and families into clinical care and promoting shared decision

making are becoming priorities throughout medical care. The study’s results can help inform

these efforts. We did not detect a difference in measures of decision making or patient

activation between the 2 groups. Both groups had high scores on scales assessing decision self-

efficacy, preparation for decision making, caregiver activation, health care satisfaction,

decisional conflict, and decisional regret. These results suggest that both groups were well

informed and well prepared to participate in their child’s treatment decision. We learned that

consultations that encourage patient participation (verbal or technology based) in their medical

care in the emergency/urgent care setting can lead to high scores on measures of decision

making. However, without baseline data for measures of decision making, we cannot know if

these types of consultations are better than usual care.

Generalizability

The results of this study should be generalizable to the larger patient population of

children with appendicitis. We had broad enrollment, with 86% of approached eligible patients

enrolling in the study. Our institution also serves as the primary surgical care facility for children

from both urban and rural communities. The high caregiver scores on measures of decision

56

making in both groups in this study suggest that caregivers can participate in a consultation that

promotes shared decision making in the emergent/urgent setting.

Subpopulation Consideration

We performed interaction tests to assess differences in the effect of the PAT based on

age, gender, race, ethnicity, and urban vs rural residence. We found no differential effects of

the PAT in any particular subgroup based on any subpopulation.

Study Limitations

This single-site study did not demonstrate that an electronic, interactive PAT could

improve decision making and patient-centered outcomes in appendicitis. We compared the PAT

to a standardized scripted surgical consultation that was developed with our stakeholders. This

script presented each treatment option with its risks and benefits and emphasized the

importance of patient value and participation in their care. The standardized consultation in

this study may have activated and encouraged patients and their families to participate in the

decision-making process about their care more than a typical consultation in clinical practice. As

we developed this standardized consultation with our stakeholders for this study, we did not

have baseline data on our outcomes using it. Upon analysis of the trial, the results

demonstrated that the scores for our outcomes measures in the standardized consultation

group were extremely high and not likely to be able to be improved on. Therefore, this study

was unable to detect an improvement with the PAT. An additional related limitation may be

due the change of having a physician member of the research team perform the enrollment

and administer the assigned intervention during the study based on stakeholder feedback. We

used research fellows to administer both interventions, which may have led to contamination

of the interventions.

The research fellows might have unknowingly included aspects of the PAT in their verbal

communication with patients and caregivers randomized to the standardized surgical

consultation alone. Therefore, the standardized consultation in this study may not be reflective

57

of current practice in other settings because it was performed by a small group of research

fellows well versed in the risks and benefits of each of the treatment choices and patient

activation strategies. This may have limited our ability to detect a beneficial effect of the PAT by

increasing measures of decision making in the control group. It is also possible that we did not

detect a difference between groups because neither intervention was effective. Because our

comparator had unknown effectiveness, this study cannot determine whether both

interventions were effective or ineffective. We are limited in our ability to evaluate if the

standardized consultation in this study is better than usual care because this study did not

include baseline measurements prior to receiving a study intervention and it did not include a

usual care control group. The choice not to use a usual care comparator group was based on

PCORI’s guidelines to try to avoid usual care comparators as well as our stakeholders’ input that

the control group needed to be standardized to ensure patients randomized to the control

group were receiving the best available standard of care. Finally, it is possible that developing

and testing a tool similar to the PAT but in a different acute disease process could demonstrate

benefit.

Future Research

Additional research is needed to determine if providing a comprehensive consultation

that promotes shared decision making can lead to strong patient and family involvement in

medical care received in the emergency/urgent care setting. This future work should utilize a

cluster randomized or pre–post intervention design to minimize contamination of the control

group. The cluster randomized design would prevent contamination; however, it would not

overcome the lack of baseline outcomes measures unless the study incorporated measuring

outcomes before and after the assigned intervention. The pre–post design would allow

assessment of baseline measurements of outcomes prior to implementation of any

intervention. Another potential solution to avoid contamination effects from one group to the

other could include developing the intervention and other aspects of the research design with

stakeholders from a different institution; this might avoid the influence of these interactions on

the execution of the study.

58

In addition, we are examining satisfaction with decision in a multisite patient choice trial

of nonoperative management of appendicitis. This will enable us to report on decisional

satisfaction in a multi-institutional cohort of patients who are being counseled by more than

100 surgeons across 10 hospitals (PCORI CER-1507-31325). Finally, this type of interactive tool

may still be valuable in the emergency department setting where the medical staff does not

always have time to explain each aspect of both treatments to the patient and his or her

caregiver. Future studies comparing this type of PAT to a less structured usual care group may

be warranted.

CONCLUSIONS Our results demonstrate that a technology-based tool did not improve measures of

decision making or patient activation for patients in the emergency/urgent care setting

compared with a standardized consultation that emphasized patient choice. Both groups had

high scores on scales assessing decision making, suggesting that both were well informed and

well prepared to participate in their child’s treatment decision. Our study suggests that a

consultation promoting shared decision making may lead to confident caregiver decision

making in the emergency/urgent care setting.

In addition, our study demonstrated the potential value of including stakeholders as

part of the research team. Changes made to our participant recruitment and retention

processes based on stakeholder input led to significant and sustained improvement in both.60

Patients, families, physicians, payers, and hospital administrators will be the key end

users of our results. Incorporating patients and families into clinical care and promoting shared

decision making are becoming priorities throughout medical care. Additional research is needed

to determine if providing a comprehensive consultation that promotes shared decision making

can lead to strong patient and family involvement in medical care in the emergency care setting

compared with usual care.

59

REFERENCES 1. US Department of Defense (DoD). Module 1: the patient activation continuum. Patient

Activation Reference Guide 2012. US DoD website. https://health.mil/Military-Health-Topics/Access-Cost-Quality-and-Safety/Quality-And-Safety-of-Healthcare/Patient-Safety/Patient-Safety-Products-And-Services/Toolkits/Patient-Activation-Reference-Guide. Accessed October 26, 2012.

2. Greene J, Hibbard JH, Tusler M. How Much Do Health Literacy and Patient Activation Contribute to Older Adults' Ability to Manage Their Health? Washington, DC: AARP Public Policy Institute; 2005. https://assets.aarp.org/rgcenter/health/2005_05_literacy.pdf

3. Hibbard JH, Cunningham PJ. How engaged are consumers in their health and health care, and why does it matter. Res Brief. 2008;8:1-9.

4. Mosen DM, Schmittdiel J, Hibbard J, Sobel D, Remmers C, Bellows J. Is patient activation associated with outcomes of care for adults with chronic conditions? J Ambul Care Manage. 2007;30(1):21-29.

5. Rask KJ, Ziemer DC, Kohler SA, Hawley JN, Arinde FJ, Barnes CS. Patient activation is associated with healthy behaviors and ease in managing diabetes in an indigent population. Diabetes Educ. 2009;35(4):622-630.

6. Remmers C, Hibbard J, Mosen DM, Wagenfield M, Hoye RE, Jones C. Is patient activation associated with future health outcomes and health care utilization among patients with diabetes? J Ambul Care Manage. 2009;32(4):320-327.

7. Skolasky RL, Green AF, Scharfstein D, Boult C, Reider L, Wegener ST. Psychometric properties of the patient activation measure among multimorbid older adults. Health Serv Res. 2011;46(2):457-478.

8. Ambulatory and Hospital Care Statistics Branch. National Hospital Ambulatory Medical Care Survey: 2009 Emergency Department Summary Tables. Hyattsville, MD: National Center for Health Statistics; 2009.

9. McCaig LF, Burt CW. National Hospital Ambulatory Medical Care Survey: 1999 emergency department summary. Adv Data. 2001;(320):1-34.

10. Coran AG, Adzick NS, Caldamone AA, Krummel TM, Laberge JM, Shamberger RC. Pediatric Surgery. 7th ed. Philadelphia, PA: Elsevier Inc; 2012.

11. Owings MF KL. Ambulatory and Inpatient Procedures in the United States, 1996. Hyattsville, MD: National Center for Health Statistics; 1998.

60

12. Wilms IM, de Hoog DE, de Visser DC, Janzing HM. Appendectomy versus antibiotic treatment for acute appendicitis. Cochrane Database Syst Rev. 2011;(11):CD008359. doi: 10.1002/14651858.CD008359.pub2

13. Bachur RG, Hennelly K, Callahan MJ, Monuteaux MC. Advanced radiologic imaging for pediatric appendicitis, 2005-2009: trends and outcomes. J Pediatr. 2012;160(6):1034-1038.

14. Abes M, Petik B, Kazil S. Nonoperative treatment of acute appendicitis in children. J Pediatr Surg. 2007;42(8):1439-1442.

15. Ein SH, Langer JC, Daneman A. Nonoperative management of pediatric ruptured appendix with inflammatory mass or abscess: presence of an appendicolith predicts recurrent appendicitis. J Pediatr Surg. 2005;40(10):1612-1615.

16. Ein SH, Shandling B. Is interval appendectomy necessary after rupture of an appendiceal mass? J Pediatr Surg. 1996;31(6):849-850.

17. Eriksson S, Granstrom L, Olander B, Wretlind B. Sensitivity of Interleukin-6 and C-reactive protein concentrations in the diagnosis of acute appendicitis. Eur J Surg. 1995;161(1):41-45.

18. Hansson J, Korner U, Khorram-Manesh A, Solberg A, Lundholm K. Randomized clinical trial of antibiotic therapy versus appendicectomy as primary treatment of acute appendicitis in unselected patients. Br J Surg. 2009;96(5):473-481.

19. Hansson J, Korner U, Ludwigs K, Johnsson E, Jonsson C, Lundholm K. Antibiotics as first-line therapy for acute appendicitis: evidence for a change in clinical practice. World J Surg. 2012;36(9):2028-2036.

20. Harrison PW. Appendicitis and the antibiotics. Am J Surg. 1953;85(2):160-163.

21. Hoffmann J, Lindhard A, Jensen HE. Appendix mass: conservative management without interval appendectomy. Am J Surg. 1984;148(3):379-382.

22. Towfigh S, Chen F, Katkhouda N, et al. Obesity should not influence the management of appendicitis. Surg Endosc. 2008;22(12):2601-2605.

23. Puapong D, Lee SL, Haigh PI, Kaminski A, Liu IL, Applebaum H. Routine interval appendectomy in children is not indicated. J Pediatr Surg. 2007;42(9):1500-1503.

24. Styrud J, Eriksson S, Nilsson I, et al. Appendectomy versus antibiotic treatment in acute appendicitis. a prospective multicenter randomized controlled trial. World J Surg. 2006;30(6):1033-1037.

61

25. Vons C, Barry C, Maitre S, et al. Amoxicillin plus clavulanic acid versus appendicectomy for treatment of acute uncomplicated appendicitis: an open-label, non-inferiority, randomised controlled trial. Lancet. 2011;377(9777):1573-1579.

26. Fitzmaurice GJ, McWilliams B, Hurreiz H, Epanomeritakis E. Antibiotics versus appendectomy in the management of acute appendicitis: a review of the current evidence. Can J Surg. 2011;54(5):307-314.

27. Malik AA, Bari SU. Conservative management of acute appendicitis. J Gastrointest Surg. 2009;13(5):966-970.

28. Shindoh J, Niwa H, Kawai K, et al. Predictive factors for negative outcomes in initial non-operative management of suspected appendicitis. J Gastrointest Surg. 2010;14(2):309-314.

29. Turhan AN, Kapan S, Kutukcu E, Yigitbas H, Hatipoglu S, Aygun E. Comparison of operative and nonoperative management of acute appendicitis. Ulus Travma Acil Cerrahi Derg. 2009;15(5):459-462.

30. Tanaka Y, Uchida H, Kawashima H, et al. Long-term outcomes of operative versus nonoperative treatment for uncomplicated appendicitis. J Pediatr Surg. 2015;50(11):1893-1897.

31. Svensson JF, Patkova B, Almstrom M, et al. Nonoperative treatment with antibiotics versus surgery for acute nonperforated appendicitis in children: a pilot randomized controlled trial. Ann Surg. 2015;261(1):67-71.

32. Salminen P, Paajanen H, Rautio T, et al. Antibiotic Therapy vs appendectomy for treatment of uncomplicated acute appendicitis: the APPAC Randomized Clinical Trial. JAMA. 2015;313(23):2340-2348.

33. Minneci PC, Mahida JB, Lodwick DL, et al. Effectiveness of patient choice in nonoperative versus surgical management of pediatric uncomplicated acute appendicitis. JAMA Surg. 2016;151(5):408-415.

34. Minneci PC, Sulkowski JP, Nacion KM, et al. Feasibility of a nonoperative management strategy for uncomplicated acute appendicitis in children. J Am Coll Surg. 2014;219(2):272-279.

35. Huang L, Yin Y, Yang L, Wang C, Li Y, Zhou Z. Comparison of antibiotic therapy and appendectomy for acute uncomplicated appendicitis in children: a meta-analysis. JAMA Pediatr. 2017;171(5):426-434.

36. Hasday SJ, Chhabra KR, Dimick JB. Antibiotics vs surgery for acute appendicitis: toward a patient-centered treatment approach. JAMA Surg. 2016;151(2):107-108.

62

37. Harnoss JC, Zelienka I, Probst P, et al. Antibiotics versus surgical therapy for uncomplicated appendicitis: systematic review and meta-analysis of controlled trials (PROSPERO 2015: CRD42015016882). Ann Surg. 2017;265(5):889-900.

38. Georgiou R, Eaton S, Stanton MP, Pierro A, Hall NJ. Efficacy and safety of nonoperative treatment for acute appendicitis: a meta-analysis. Pediatrics. 2017;139(3). doi: 10.1542/peds.2016-3003

39. Anderson LA, Sharpe PA. Improving patient and provider communication: a synthesis and review of communication interventions. Patient Educ Couns. 1991;17(2):99-134.

40. Harrington J, Noble LM, Newman SP. Improving patients' communication with doctors: a systematic review of intervention studies. Patient Educ Couns. 2004;52(1):7-16.

41. Haywood K, Marshall S, Fitzpatrick R. Patient participation in the consultation process: a structured review of intervention strategies. Patient Educ Couns. 2006;63(1-2):12-23.

42. Post DM, Cegala DJ, Miser WF. The other half of the whole: teaching patients to communicate with physicians. Fam Med. 2002;34(5):344-352.

43. Stacey D, Bennett CL, Barry MJ, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2011;(10):CD001431. doi: 10.1002/14651858.CD001431.pub3

44. Harrington NG, Norling GR, Witte FM, Taylor J, Andrews JE. The effects of communication skills training on pediatricians‘ and parents‘ communication during “sick child" visits. Health Commun. 2007;21(2):105-114.

45. Lewis CC, Pantell RH, Sharp L. Increasing patient knowledge, satisfaction, and involvement: randomized trial of a communication intervention. Pediatrics. 1991;88(2):351-358.

46. Coyne I, Gallagher P. Participation in communication and decision making: children and young people‘s experiences in a hospital setting. J Clin Nurs. 2011;20(15-16):2334-2343.

47. Jackson C, Cheater FM, Reid I. A systematic review of decision support needs of parents making child health decisions. Health Expect. 2008;11(3):232-251.

48. Lipstein EA, Brinkman WB, Britto MT. What is known about parents' treatment decisions? A narrative review of pediatric decision making. Med Decis Making. 2012;32(2):246-258.

49. Moore L, Kirk S. A literature review of children’s and young people's participation in decisions relating to health care. J Clin Nurs. 2010;19(15-16):2215-2225.

50. Pennarola BW, Rodday AM, Mayer DK, et al. Factors associated with parental activation in pediatric hematopoietic stem cell transplant. Med Care Res Rev. 2012;69(2):194-214.

63

51. Chappuy H, Taupin P, Dimet J, et al. Do parents understand the medical information provided in paediatric emergency departments? A prospective multicenter study. Acta Paediatr. 2012;101(10):1089-1094.

52. Horowitz L, Kassam-Adams N, Bergstein J. Mental health aspects of emergency medical services for children: summary of a consensus conference. J Pediatr Psychol. 2001;26(8):491-502.

53. Kassam-Adams N, Fleisher CL, Winston FK. Acute stress disorder and posttraumatic stress disorder in parents of injured children. J Trauma Stress. 2009;22(4):294-302.

54. Katan M, Christ-Crain M. The stress hormone copeptin: a new prognostic biomarker in acute illness. Swiss Med Wkly. 2010;140:w13101. doi: 10.4414/smw.2010.13101

55. Schaeffer MH, Krantz DS, Wichman A, Masur H, Reed E, Vinicky JK. The impact of disease severity on the informed consent process in clinical research. Am J Med. 1996;100(3):261-268.

56. Smithline HA, Mader TJ, Crenshaw BJ. Do patients with acute medical conditions have the capacity to give informed consent for emergency medicine research? Acad Emerg Med. 1999;6(8):776-780.

57. Wray J, Lee K, Dearmun N, Franck L. Parental anxiety and stress during children's hospitalisation: the StayClose study. J Child Health Care. 2011;15(3):163-174.

58. Hibbard JH, Stockard J, Mahoney ER, Tusler M. Development of the Patient Activation Measure (PAM): conceptualizing and measuring activation in patients and consumers. Health Serv Res. 2004;39(4, pt 1):1005-1026.

59. Ottawa Decision Support Framework. Patient decision aids. Ottawa Hospital Research Institute website. http://decisionaid.ohri.ca/odsf.html. Published 2012. Accessed August 15, 2012.

60. Minneci PC, Nacion KM, Lodwick DL, Cooper JN, Deans KJ. Improving surgical research by involving stakeholders. JAMA Surg. 2016;151(6):579-580.

61. Greenfield S, Kaplan S, Ware JE Jr. Expanding patient involvement in care. Effects on patient outcomes. Ann Intern Med. 1985;102(4):520-528.

62. Cegala DJ, McClure L, Marinelli TM, Post DM. The effects of communication skills training on patients' participation during medical interviews. Patient Educ Couns. 2000;41(2):209-222.

63. Cegala DJ, Post DM, McClure L. The effects of patient communication skills training on the discourse of older patients during a primary care interview. J Am Geriatr Soc. 2001;49(11):1505-1511.

64

64. Cegala DJ, Street RL Jr, Clinch CR. The impact of patient participation on physicians' information provision during a primary care medical interview. Health Commun. 2007;21(2):177-185.

65. Cegala DJ, Post DM. The impact of patients' participation on physicians' patient-centered communication. Patient Educ Couns. 2009;77(2):202-208.

66. Bunn H, O'Connor A. Validation of client decision-making instruments in the context of psychiatry. Can J Nurs Res. 1996;28(3):13-27.

67. Graham I, O'Connor A. User manual-preparation for decision making scale. http://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_PrepDM.pdf. Published 1995. Accessed December 1, 2012.

68. Cranney A, O'Connor AM, Jacobsen MJ, et al. Development and pilot testing of a decision aid for postmenopausal women with osteoporosis. Patient Educ Couns. 2002;47(3):245-255.

69. Varni JW, Burwinkle TM, Dickinson P, et al. Evaluation of the built environment at a children's convalescent hospital: development of the Pediatric Quality of Life Inventory parent and staff satisfaction measures for pediatric health care facilities. J Dev Behav Pediatr. 2004;25(1):10-20.

70. Varni JW, Quiggins DJL, Ayala GX. Development of the pediatric hematology/oncology parent satisfaction survey. Child Health Care. 2000;29(4):243-255.

71. Bennett C, Graham ID, Kristjansson E, Kearing SA, Clay KF, O'Connor AM. Validation of a preparation for decision making scale. Patient Educ Couns. 2010;78(1):130-133.

72. O'Connor A, Jacobsen MJ, Elmslie T, et al. Simple vs complex patient decision aids: is more necessarily better? Med Decis Making. 2000;20(4):496.

73. Stacey D, O'Connor AM, DeGrasse C, Verma S. Development and evaluation of a breast cancer prevention decision aid for higher-risk women. Health Expect. 2003;6(1):3-18.

74. O'Connor A. User manual—decisional conflict scale (10-item question format). http://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Decisional_Conflict.pdf. Published 1993. Accessed August 13, 2012.

75. O'Connor AM. Validation of a decisional conflict scale. Med Decis Making. 1995;15(1):25-30.

76. Linder SK, Swank PR, Vernon SW, Mullen PD, Morgan RO, Volk RJ. Validity of a low literacy version of the Decisional Conflict Scale. Patient Educ Couns. 2011;85(3):521-524.

65

77. Jibaja-Weiss ML, Volk RJ, Friedman LC, et al. Preliminary testing of a just-in-time, user-defined values clarification exercise to aid lower literate women in making informed breast cancer treatment decisions. Health Expect. 2006;9(3):218-231.

78. Volk RJ, Jibaja-Weiss ML, Hawley ST, et al. Entertainment education for prostate cancer screening: a randomized trial among primary care patients with low health literacy. Patient Educ Couns. 2008;73(3):482-489.

79. Bunn H, Lange I, Urrutia M, et al. Health preferences and decision-making needs of disadvantaged women. J Adv Nurs. 2006;56(3):247-260.

80. O'Connor AM. User manual—Decision Regret Scale. http://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Regret_Scale.pdf. Published 1996. Accessed August 13, 2012.

81. Brehaut JC, O‘Connor AM, Wood TJ, et al. Validation of a decision regret scale. Med Decis Making. 2003;23(4):281-292.

82. Davison BJ, Goldenberg SL. Decisional regret and quality of life after participating in medical decision-making for early-stage prostate cancer. BJU Int. 2003;91(1):14-17.

83. Hickman RL Jr, Pinto MD, Lee E, Daly BJ. Exploratory and confirmatory factor analysis of the decision regret scale in recipients of internal cardioverter defibrillators. J Nurs Meas. 2012;20(1):21-34.

84. Varni JW. Pediatric Quality of Life Inventory (PedsQLTM). http://www.pedsql.org/. Published 2012. Accessed August 15, 2012.

85. Schurman JV, Cushing CC, Garey CL, Laituri CA, St Peter SD. Quality of life assessment between laparoscopic appendectomy at presentation and interval appendectomy for perforated appendicitis with abscess: analysis of a prospective randomized trial. J Pediatr Surg. 2011;46(6):1121-1125.

86. Varni JW, Burwinkle TM, Seid M, Skarr D. The PedsQL 4.0 as a pediatric population health measure: feasibility, reliability, and validity. Ambul Pediatr. 2003;3(6):329-341.

87. Varni JW, Limbers CA, Burwinkle TM. Parent proxy-report of their children's health-related quality of life: an analysis of 13,878 parents' reliability and validity across age subgroups using the PedsQL 4.0 Generic Core Scales. Health Qual Life Outcomes. 2007;5:2.

88. Varni JW, Seid M, Kurtin PS. PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory version 4.0 generic core scales in healthy and patient populations. Med Care. 2001;39(8):800-812.

66

89. Varni JW, Limbers CA, Newman DA. Using factor analysis to confirm the validity of children's self-reported health-related quality of life across different modes of administration. Clin Trials. 2009;6(2):185-195.

90. Limbers CA, Newman DA, Varni JW. Factorial invariance of child self-report across socioeconomic status groups: a multigroup confirmatory factor analysis utilizing the PedsQL 4.0 Generic Core Scales. J Behav Med. 2008;31(5):401-411.

91. Varni JW, Limbers CA. The PedsQL 4.0 Generic core scales young adult version: feasibility, reliability, and validity in a university student population. J Health Psychol. 2009;14(4):611-622.

92. Fraenkel L, Rabidou N, Wittink D, Fried T. Improving informed decision making for patients with knee pain. J Rheumatol. 2007;34(9):1894-1898.

93. Zlowodzki M, Bhandari M. Outcome measures and implications for sample-size calculations. J Bone Joint Surg Am. 2009;3(suppl 91):35-40.

94. Mason RJ, Moazzez A, Sohn H, Katkhouda N. Meta-analysis of randomized trials comparing antibiotic therapy with appendectomy for acute uncomplicated (no abscess or phlegmon) appendicitis. Surg Infect (Larchmt). 2012;13(2):74-84.

95. Wiens BL, Dmitrienko A. The fallback procedure for evaluating a single family of hypotheses. J Biopharm Stat. 2005;15(6):929-942.

96. Morikawa T, Yoshida M. A useful testing strategy in phase III trials: combined test of superiority and test of equivalence. J Biopharm Stat. 1995;5(3):297-306.

97. Committee for Propietary Medicinal Products (CPMP), European Agency for the Evaluation of Medicinal Products (EMEA). Points to Consider on Switching Between Superiority and Noninferiority. Report CPMP/EWP/482/99. London, UK: EMEA; 2000. https://www.ema.europa.eu/en/documents/scientific-guideline/points-consider-switching-between-superiority-non-inferiority_en.pdf

98. Dunnett CW, Tamhane AC. Multiple testing to establish superiority/equivalence of a new treatment compared with kappa standard treatments. Stat Med. 1997;16(21):2489-2506.

99. Ke C, Ding B, Jiang Q, Snapinn SM. The issue of multiplicity in noninferiority studies. Clin Trials. 2012;9(6):730-735.

100. National Research Council. The Prevention and Treatment of Missing Data in Clinical Trials. Washington, DC: National Academies Press; 2010.

67

PUBLICATIONS Minneci PC, Sulkowski JP, Nacion KM, et al. Feasibility of a nonoperative management strategy for uncomplicated acute appendicitis in children. J Am Coll Surg. 2014; 219(2):272-279. doi: 10.1016/j.jamcollsurg.2014.02.031

Minneci PC, Deans KJ. Treatment of uncomplicated acute appendicitis. JAMA. 2015; 314(13):1401-1402. doi: 10.1001/jama.2015.11426

Minneci PC, Mahida JB, Lodwick DL, et al. Effectiveness of patient choice in nonoperative vs surgical management of pediatric uncomplicated acute appendicitis. JAMA Surg. 2016;151(5):408-415. doi: 10.1001/jamasurg.2015.4534

Minneci PC, Nacion KM, Lodwick DL, Cooper JN, Deans KJ. Improving surgical research by involving stakeholders. JAMA Surg. 2016;151(6):579-580. doi:10.1001/jamasurg.2015.4898

Minneci PC, Deans KJ. Management of acute appendicitis, comparative effectiveness research, and the nuances of study design-reply. JAMA Surg. 2016;151(8):784-785. doi: 10.1001/jamasurg.2016.0478

Gonzalez DO, Deans KJ, Minneci PC. Role of non-operative management in pediatric appendicitis. Semin Pediatr Surg. 2016;25(4):204-207.doi:10.1053/j.sempedsurg.2016.05.002

Lopez JJ, Deans KJ, Minneci PC. Nonoperative management of appendicitis in children. Curr Opin Pediatr. 2017;29(3):358-362. doi: 10.1097/MOP.0000000000000487

Minneci PC, Deans KJ. Is an RCT the best way to investigate the effectiveness of nonoperative management of pediatric appendicitis? Ann Surg. 2017;266(1):e5-e6. doi: 10.1097/SLA.0000000000001275

Radio and Television Minneci PC, Deans KJ. Appendicitis. WXDE, Lewes, DE (DMA#143). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis. WHAM-AM, Rochester, NY (DMA#78). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis. WTVN-AM, Columbus, OH (DMA#32). December 16, 2015.

Minneci PC, Deans KJ. New study could change the way doctors treat appendicitis. NBC Nightly News. December 15, 2015.

http://www.nbcnews.com/nightly-news/video/new-study-could-change-the-way-doctors-treat-appendicitis-587261507815.

Minneci PC, Deans KJ. Appendicitis treatment. Channel 6 News. December 15, 2015.

68

https://www.iqmediacorp.com/ClipPlayer/?ClipID=961829ba-af1d-4bc3-8a5f-674e93882eb7.

Minneci PC, Deans KJ. Appendicitis. KTRH-AM, Houston, TX (DMA#10). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis Treatment. WBZ-AM, Boston, MA (DMA#7). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis. WWJ-AM, Detroit, MI (DMA#12). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis Treatment. WOR-AM, New York, NY (DMA#1). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis. WCBS-NY, New York, NY (DMA#1). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis. WBBM-AM, Chicago, IL (DMA#3). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis. KRLD-AM, Dallas, TX (DMA#5). December 16, 2015.

Minneci PC, Deans KJ. Appendicitis Treatment” WMEX-AM, Boston, MA (DMA#7). December 16, 2015.

Multimedia/Database/Website Seaman AM. Appendicitis without surgery may be safe option for some kids. Thomson Reuters Health News. https://www.reuters.com/article/us-appendicitis-surgery/appendicitis-without-surgery-may-be-safe-option-for-some-kids-idUSBREA3L1FR20140422. Published April 22, 2014.

Bernstein L. Antibiotics instead of surgery for appendicitis in children. Washington Post. https://www.washingtonpost.com/news/to-your-health/wp/2014/04/28/antibiotics-instead-of-surgery-for-appendicitis-in-children/?utm_term=.6d91a8e4d934. Published April 28, 2014.

Deans KJ, Minneci PC. Treating appendicitis without surgery. New York Times. April 25, 2015.

Minneci PC, Deans KJ. Engaging patients and other stakeholders in PCORI research projects. PCORI website. Published November 20, 2015.

Landro L. New tools help patients make tough decisions in the ER. The Wall Street Journal. https://www.wsj.com/articles/new-tools-help-patients-make-tough-decisions-in-the-er-1461608262. Published April 25, 2016.

Deans KJ, Minneci PC. Antibiotics could replace some appendectomies. ThisWeek Newspaper. September 2014.

Minneci PC. Treating appendicitis with antibiotics. Columbus Parent. https://www.columbusparent.com/content/stories/2014/09/pediatric-healthsource.html. Published August 25, 2014.

69

Minneci PC. Decision making matters in appendicitis. Physician’s Weekly. https://www.physiciansweekly.com/decision-making-matters-in-appendicitis/. Published October 25, 2016.

Brazier Y. Patient involvement: what is it, and why does it matter? http://www.medicalnewstoday.com/articles/305702.php. Published January 28, 2016.

Whitlock Burton K. Antibiotics alone are a successful treatment for uncomplicated acute appendicitis in children. Pediatrics Nationwide blog. https://pediatricsnationwide.org/2014/05/14/antibiotics-alone-are-a-successful-treatment-for-uncomplicated-acute-appendicitis-in-children/. Published May 14, 2014.

DiMartini A. Involving stakeholders boosts clinical trial participation. Pediatrics Nationwide blog. http://pediatricsnationwide.org/2016/02/10/involving-stakeholders-boosts-clinical-trial-participation/?utm_campaign=shareaholic&utm_medium=email_this&utm_source=email. Published February 10, 2016.

70

APPENDIX Standardized Surgical Consultation Script:

Appendicitis is inflammation/infection of your appendix.

Since your child’s appendicitis was diagnosed early, there are 2 treatment options: surgery or nonoperative treatment with antibiotics alone. These 2 treatment options are both reasonable choices but they have very different risks and benefits. We will review what each treatment involves and then the associated risks and benefits of each.

The first option is an appendectomy which is surgery to remove your child’s appendix. If you choose this option, your child will be admitted and given IV antibiotics up until surgery. For surgery, your child will be given general anesthesia to put him/her to sleep. Surgery is performed by making 3 small incisions and using a camera and 2 instruments to find and remove the appendix. The incisions are closed and your child is awoken from anesthesia. The entire procedure takes about one hour. After surgery, patients usually stay in the hospital one to 2 days. Your child can usually return to school in few days and resume all activities, including sports, in about 2 weeks.

The second option is to treat your child’s appendicitis with antibiotics alone. If you choose this option, your child will be admitted and given IV antibiotics for at least one day. Usually, with antibiotics alone, patients stay in the hospital 1 to 2 days. Doctors and nurses will check on your child frequently to make sure he/she is getting better. When your child feels better and is able to eat, he or she will continue taking the antibiotics by mouth for about 7 days at home. Children can typically return to all activities, including school and sports, in 2 to 3 days.

Either of these treatment options is reasonable and there is no right or wrong choice. However, there is likely one choice that is best for you and your child depending on which of the risks and benefits are most important to you.

In order to help you make the best decision for your child and family, let’s look at the risks and benefits of each treatment option.

The benefits of nonoperative treatment with antibiotics alone include:

- Antibiotics alone have been shown to be a safe method to treat children with acute appendicitis. About 8 of 10 children never need surgery

- Your child’s pain may go away faster and he/she will recover sooner. - If your child never needs surgery, then there are no risks of surgery.

The possible risks of nonoperative treatment may include:

71

- Your child’s symptoms might not go away and he/she will need an appendectomy which involves the risks of surgery. About 1 of 10 children do not get better and will need surgery while in the hospital

- Your child’s appendicitis could come back in the future. About 1 out of 10 children will have appendicitis again.

- Altogether, about 2 out 10 patients treated with antibiotics alone will eventually need an appendectomy.

- There can be side effects of antibiotics o Most common: nausea, vomiting and diarrhea

The benefits of surgery include:

- Surgery is curative. Your child will never have appendicitis again. - Surgery is the most common way to treat appendicitis. About 9 out of 10 children will

not have a complication after appendectomy. - Your child can usually go home within 1 to 2 days after surgery

The possible risks of surgery may include:

- Your child will be in some pain after surgery - Most kids need a few days of rest before going back to school and 1-2 weeks before

resuming physical activity - It will leave 1-3 small scars on your belly - There are some risks during surgery, such as bleeding or problems from the anesthesia.

About 1 out of 10 patients experience a complication. - The most common complications are minor and include infections or problems with the

wounds. - Other possible risks also include:

o Developing an abdominal abscess (an infection inside your child’s belly) o Scars (adhesions) that can cause future blockage in your child’s belly that may

require additional surgery - If your child has a complication, then his/her hospital stay could be extended and

he/she may need more medications such as antibiotics

Both surgery and nonoperative treatment with antibiotics alone are good treatments for your child’s appendicitis. Either of these treatment options is reasonable and there is no right or wrong choice. You should choose whichever treatment is best for your child and family based on which risks and benefits are most important to you.

PAT Video demonstration: https://vimeo.com/91207174

72

Copyright© 2020. The Research Institute at Nationwide Children's Hospital. All Rights Reserved.

Disclaimer: The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee.

Acknowledgement:

Research reported in this report was [partially] funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (#CE-12-11-4350) Further information available at: https://www.pcori.org/research-results/2013/helping-caregivers-and-their-children-early-appendicitis-make-treatment