predictors of response to physical therapy intervention for · predictors of response to physical...

For Peer Review

Predictors of Response to Physical Therapy Intervention for

Plantar Heel Pain: A Secondary Analysis of Data from a

Randomized Clinical Trial.

Journal: Foot & Ankle International

Manuscript ID: FAI-14-0438.R2

Manuscript Type: (Original) Clinical Research Article

Keywords: Exercise, Manual Therapy, Obese, Plantar Fasciitis, Prognosis

STRUCTURED ABSTRACT (Limit 250 words) <br>

Background:

Background: Age, weight, and duration of symptoms have been associated

with a poor response to treatment for plantar heel pain (PHP), but have not been examined in response to physical therapy (PT) intervention. The purpose of this investigation was to examine the influence of age, body mass index (BMI), and symptom duration on treatment response to PT intervention. Methods: Sixty participants received 6 sessions of PT intervention over 4 weeks that included manual therapy and exercise or electrophysiological agents and exercise. Outcomes were assessed at 6 months using the Foot and Ankle Ability Measure (FAAM), Numeric Pain Rating Scale (NPRS), and Global Rating of Change Scale (GRC). Logistic regression was used to analyze age, BMI, and symptom duration as predictors of a successful response based on minimal clinically important changes in the FAAM, NPRS

and GRC, or only the FAAM and NPRS. Results: Improvement in NPRS was 3 points (95% CI; 2.4, 3.6) and FAAM improved by 22.5 points (95% CI; 16.8, 28.2). Individuals with symptoms less than 7.2 months were 4.2 (95% CI; 1.3, 13.8; p=.016) and 8.5 (95% CI; 2.5, 28.9; p=.001) times more likely to respond based on the NPRS/FAAM/GRC, and NPRS/FAAM success criteria, respectively. Age and BMI were not significant predictors. Conclusion: Body mass was not associated with outcomes and obese individuals can achieve clinical success with the PT intervention used in the clinical trial. Individuals with PHP symptoms longer than 7 months require additional consideration and further investigation of effective strategies to improve treatment response.

http://mc.manuscriptcentral.com/fai

Foot & Ankle International

For Peer Review

Predictors of Response to Physical Therapy Intervention for Plantar Heel 1

Pain: A Secondary Analysis of Data from a Randomized Clinical Trial. 2

3

4

Abstract 5

Background: Age, weight, and duration of symptoms have been associated with 6

a poor response to treatment for plantar heel pain (PHP), but no studies were 7

identified that examined predictors of response to physical therapy intervention. 8

The purpose of this investigation was to examine the influence of age, body 9

mass index (BMI), and symptom duration on treatment response to physical 10

therapy intervention. 11

Methods: Sixty participants received 6 visits over 4 weeks of physical therapy 12

intervention that included manual therapy and exercise or electrophysiological 13

agents and exercise. Outcomes were assessed using the Foot and Ankle Ability 14

Measure (FAAM), Numeric Pain Rating Scale (NPRS), and Global Rating of 15

Change Scale (GRC). Logistic regression (p<.05) was used to analyze age, BMI, 16

and symptom duration as potential predictors of a successful response based on 17

the minimal clinically important difference of the outcome measures. Sensitivity 18

analysis was used to assess the influence of success based on minimal clinically 19

important changes in the FAAM, NPRS and GRC, or only the FAAM and NPRS. 20

Receiver operating curves were used to determine the cut point for the significant 21

predictor. 22

Results: At the 6 month follow-up to physical therapy intervention, NPRS was 23

improved by 3 points (95% CI; 2.4, 3.6) and FAAM improved by 22.5 points (95% 24

CI; 16.8, 28.2). Individuals with symptoms less than 7.2 months were 4.2 (95% 25

CI; 1.3, 13.8; p=.016) and 8.5 (95% CI; 2.5, 28.9; p=.001) times more likely to 26

Page 1 of 25



For Peer Review

respond to treatment based on the NPRS/FAAM/GRC, and NPRS/FAAM 27

success criteria, respectively. Age and BMI were not significant predictors (p ≥ 28

.455 and p ≥ .450, respectively). 29

Conclusion: Age and BMI were not associated with outcomes and obese 30

individuals did achieve a successful outcome with the physical therapy 31

intervention used in the clinical trial. Individuals with PHP symptoms longer than 32

7 months require additional consideration and further investigation of effective 33

strategies to improve treatment response. 34

Level of Evidence: Prognosis, level 2b 35

36

Key Words: Plantar Fasciitis, Prognosis, Obese, Manual Therapy, Exercise. 37

38

39

Introduction 40

41

Plantar heel pain (PHP), commonly referred to as plantar fasciitis, is a 42

frequently occurring foot condition that results in disability and limited function in 43

work, recreation, and daily activities. While plantar fasciitis has been reported as 44

the most common cause of plantar heel pain, the term plantar heel pain is 45

inclusive of other pathological and painful conditions of the plantar fascia and 46

heel that are often difficult to discriminate via history and physical examination.6, 47

68 Individuals with PHP incur approximately 800,000 to 1 million visits to 48

physicians annually at an estimated cost of 192 to 376 million US dollars.55, 62 In 49

addition to physician visits, PHP was the most prevalent condition treated by 50

podiatrists and the most common foot condition seen by physical therapists.2, 52 51

Conservative (ie, non-surgical and non-extracorporeal shock wave therapy) 52

Page 2 of 25



For Peer Review

interventions are advocated initially to manage PHP and most patients respond 53

positively to treatment.13, 42, 48, 61 Despite improvement in most who are treated 54

conservatively for PHP, approximately 18-50% of individuals continue to have 55

symptoms after conservative treatment and 30% have recurrent symptoms.5, 15, 56

39, 67 57

Incomplete recovery or failure to respond to conservative treatment may 58

be attributed to patient characteristics that affect treatment response. Age, body 59

mass, and duration of symptoms have been identified as characteristics 60

associated with a reduced response to conservative treatment for plantar heel 61

pain (PHP).21, 30, 39, 51, 67 Recently, a randomized clinical trial demonstrated 62

clinically meaningful improvement following multi-modal physical therapy 63

intervention for PHP,8 but did not analyze the influence of age, body mass, or 64

duration of symptoms on treatment response. No other evidence was found to 65

indicate factors associated with treatment response to multi-modal physical 66

therapy intervention. Knowledge of factors that can potentially affect outcome 67

can help identify individuals less likely to respond to physical therapy 68

intervention. Once identified, additional research can focus on strategies that 69

may be more effective for subgroups of nonresponders and help to reduce the 70

incidence of persistent or recurrent symptoms in individuals with PHP. Therefore, 71

the purpose of this investigation was to identify if age, body mass, or symptom 72

duration were predictors of outcome to physical therapy intervention in 73

individuals with PHP. 74

75

76

Materials and Methods 77

78

Page 3 of 25



For Peer Review

Participants 79

The study population included 60 individuals who participated in a 80

randomized clinical trial between October 2006 and January 2008.8 Participants 81

included patients with PHP between the ages of 27 and 63 years that presented 82

to physical therapy at 1 of 2 outpatient orthopaedic physical therapy clinics. The 83

diagnosis of PHP was based on pain locatedalized to at the medical calcaneal 84

tubercle and pain worstened withduring the first few steps in the moringafter 85

waking. clinical presentation without additional iImaging was not used to further 86

discriminate the pathological conditions associated with the clinical presentation 87

of PHP.studies. Patients with precautions to manual therapy interventions (ie, 88

tumor, fracture, rheumatoid arthritis, osteoporosis, prolonged history of steroid 89

use, severe vascular disease, etc), prior surgery to the distal tibia, fibula, ankle 90

joint, or rearfoot region (proximal to the base of the metatarsals), insufficient 91

English proficiency to complete questionnaires, or inability to comply with 92

treatment and follow-up schedules were excluded. Prior to participating in the 93

clinical trial, patients reviewed and signed a consent form approved by The 94

Human Investigations Committee, XXXXXX and the XXXXXX prior to 95

participation in the clinical trial.8 96

97

Measures 98

All participants in the clinical trial completed a standardized baseline 99

examination at the outset that recorded age, sex, duration of symptoms, body 100

mass index (BMI), and use of medications. In addition, the activities of daily living 101

subscale of the Foot and Ankle Ability Measure (FAAM), the Lower Extremity 102

Functional Scale (LEFS), the Beck Anxiety Index (BAI), and the 3-item (current, 103

best, worst pain) Numeric Pain Rating Scale (NPRS) were completed at 104

Page 4 of 25



For Peer Review

baseline, 4 weeks, and 6 months after initiation of treatment. Also, the Global 105

Rating of Change Scale (GRC) was completed at 4 weeks and 6 months after 106

baseline. For the purpose of this secondary analysis, the variables selected 107

included age, duration of symptoms, BMI, FAAM, NPRS, and GRC. The FAAM 108

consists of 21 questions about the difficulty (range; 0 = unable to do, through 4 = 109

no difficulty) of daily tasks due to foot and ankle problems. The total score is 110

divided by the highest possible score and multiplied by 100, where 0 = no 111

functional ability and 100 = highest state of functional ability. The FAAM has 112

demonstrated content and context validity, an intraclass correlation coefficient 113

(ICC) of 0.89, and a minimal clinically important difference (MCID) of 8 points.38 114

An 11-point NPRS, where 0 = no pain and 10 = worst imaginable pain was used 115

to capture pain intensity. The NPRS with 3-items has demonstrated validity, a 116

test-retest reliability of 0.61 to 0.88 and a MCID of 2 points.7, 18, 31, 32 The GRC is 117

a single item questionnaire that ranks patient-perceived improvement between -7 118

(a very great deal worse) to 0 (about the same) to +7 (a very great deal better).29 119

Scores of +5 (quite a bit better) have been used as an indicator of clinical 120

success.9, 29, 35 121

122

Procedures 123

Complete details of the clinical trial procedures are described by Cleland 124

et al.8 Participants received 6 visits over 4 weeks of physical therapy intervention 125

that included manual therapy and exercise or electrophysiological agents and 126

exercise. Both groups received exercises that included stretching and self-127

mobilization of the plantar fascia, calf stretching with the knee flexed and 128

extended, and ankle eversion self-mobilization. All patients were instructed to 129

perform all activities of daily living that did not increase symptoms and to avoid 130

Page 5 of 25



For Peer Review

aggravating activities. In the group that received manual therapy, intervention 131

was based upon patient-specific impairments in the hip, knee, lower leg, ankle, 132

and foot regions. Manual therapy included joint mobilization or manipulation to 133

the aforementioned areas and soft tissue mobilization to the plantar fascia and 134

flexor hallicus longus. In the group that received electrophysiological agents, 135

ultrasound (3 MHz, 1.5 W/cm2, 100-Hz, 20% duty cycle, 5 minute duration) was 136

performed to the involved area followed by iontophoresis with dexamethasone 137

(40 mA·min dose). Thirteen patients were taking anti-inflammatory medication at 138

the start of the study and were instructed to continue usage as prescribed by 139

their physician. Patients were limited to intervention provided within the clinical 140

trial and did not receive any other intervention; eg, injections, night splints, anti-141

inflammatory medication (other than that taken prior to outset of the study), or 142

immobilization were not utilized. Both groups were collapsed into one to 143

determine response to physical therapy intervention that included exercise and 144

modalities or exercise and manual therapy. The FAAM, LEFS, BAI, NPRS, and 145

GRC were administered at 4 weeks and 6 months after the start of treatment. 146

147

Statistical Analysis 148

All data analyses and imputations were performed using SPSS, Version 149

22.0 (SPSS Inc., Chicago, IL, USA). Despite significant difference in outcomes 150

between the 2 physical therapy intervention groups, significant, and clinically 151

meaningful, improvement was observed in both groups.8 Therefore, to provide 152

the most comprehensive analysis that included all responders to physical therapy 153

intervention, participants from both treatment groups were included in the 154

analysis consistent with multivariate prognostic modelling recommendations and 155

procedures used by other responder analyses of randomized clinical trials.20, 24, 156

Page 6 of 25



For Peer Review

25, 33, 46 This type of analysis allows a detailed analysis of whether age, body 157

mass, or symptom duration should alter the decision to manage PHP with 158

physical therapy interventions. Less than 12% of the original data was missing 159

for the variables used in this investigation. Multiple imputation procedures were 160

used to handle missing values after analysis of the missing data and observed 161

absence of monotonicity. Five imputed data sets were generated using the 162

Markov Chain Monte Carlo method and pooled estimates from the multiple 163

imputations were used when reporting data. Patient characteristics and outcome 164

variables were summarized using the mean and standard deviation for 165

continuous measures and the frequency and percentages for categorical 166

measures (Table 1). 167

Standard logistic regression (p < .05) was used to analyze age, BMI, and 168

symptom duration as predictors of successful treatment response based on the 169

minimal clinically important difference (MCID) of the outcome measures. 170

Predictor variables were tested for multicollinearity and demonstrated variance 171

inflation factor values greater than 10 and tolerance values less than 0.10 in all 172

imputed data sets. A sensitivity analysis was conducted using 2 criteria to define 173

successful response to physical therapy intervention. In the first scenario, 174

participants were defined as having a successful response if they exceeded 175

MCID thresholds for 3 outcome measures (NPRS, 2 points; FAAM, 8 points; and 176

GRC, +5 or greater) at the 6 month follow-up similar to procedures used by 177

Cleland et al.10 In the second scenario, successful response was defined only 178

using the MCID of the NPRS and FAAM. Recent evidence has demonstrated 179

concerns regarding the validity and stability of the GRC and therefore both 180

analyses were conducted to assess the influence of a successful response 181

criterion that does not include the GRC.22, 57 Receiver operating curves were 182

Page 7 of 25



For Peer Review

used to determine the cut point for the significant predictor and the Nagelkerke 183

R2 was used to describe the amount of variation explained by the significant 184

predictor. The odds ratio (95% CI) was reported to indicate the odds of a 185

successful response to treatment based on the determined cut point of the 186

significant predictor. Sensitivity, specificity, and positive and negative likelihood 187

ratio values were calculated for the predictor variable. In addition, MANCOVA 188

and Chi-square tests were used to compare the FAAM, NPRS, and GRC 189

between individuals above and below the cut point for the significant predictor 190

using baseline scores as the covariate. 191

Separate post hoc power analyses for logistic regression analyses were 192

performed using G*Power 3.1.6 using the observed odds ratio, Pr (y = 1 | x = 1) 193

H0 calculated from classification tables and R2 obtained from both response 194

criteria analyses, and α = .05.19 195

196

197

Results 198

199

The characteristics of the sample included middle-aged participants who 200

were, on average, obese (Table 1).45 Mean changes in NPRS and FAAM scores 201

in response to physical therapy intervention at the 6 month follow-up exceeded 202

the MCID of 2 and 8 points, respectively (Table 1). Also, the lower confidence 203

interval of the FAAM and NPRS change score exceeded the MCID. In addition, 204

67% of participants reported a GRC of +5 (quite a bit better) or greater. 205

The sensitivity analysis demonstrated different frequencies of successful 206

response between the 2 criteria used. The NPRS/FAAM/GRC criteria resulted in 207

29 cases deemed a successful response and the NPRS/FAAM criteria defined a 208

Page 8 of 25



For Peer Review

successful response in 38 cases (Table 1). Duration of symptoms was the only 209

significant predictor in logistic regression analyses using the NPRS/FAAM/GRC 210

and the NPRS/FAAM criteria. Age and BMI were not significant predictors (p ≥ 211

.455 and p ≥ .450, respectively) of successful treatment response. Analysis of the 212

receiver operating curves for symptom duration identified a cut-point of 218 days 213

(ie, 7.2 months) for both successful response criteria. Individuals with symptoms 214

less than 7.2 months were 4.2 (p = .016) and 8.5 (p = .001) times more likely to 215

respond to treatment based on the NPRS/FAAM/GRC, and NPRS/FAAM 216

success criteria, respectively (Table 2). Sensitivity, specificity, and positive and 217

negative likelihood ratio values for the accuracy of symptoms less than 7.2 218

months to predict a positive response to physical therapy intervention are listed 219

in Table 2. Using the response rate of the sample and calculated likelihood ratios 220

(Tables 1 and 2), the probability of successful response to physical therapy 221

intervention if symptoms were present less than 7.2 months increased from 48% 222

to 68% (95% CI 52, 81) using the NPRS/FAAM/GRC criteria and from 63% to 223

83% (95% CI 70, 91) using the NPRS/FAAM criteria. Differences in the FAAM, 224

NPRS, and GRC between individuals with symptoms less than, and greater than, 225

7.2 months are provided in Table 3. 226

For the logistic regression analysis using the NPRS/FAAM/GRC criteria, 227

the Nagelkerke R2 ranged from 0.117 to 0.181 in the 5 imputed data sets and 228

indicated that duration of symptoms predicted 18.1% of the variation, at best. The 229

Nagelkerke R2 using the NPRS/FAAM criteria range between 0.258 to 0.326 and 230

therefore duration of symptoms explained 32.6% of the variation, at best. 231

Results of post-hoc power analysis indicated a power of 0.97 for the 232

analysis including the NPRS/FAAM/GRC criteria and a power of 0.99 using the 233

NPRS/FAAM criteria. 234

Page 9 of 25



For Peer Review

235

236

Discussion 237

In the clinical trial, 48% to 63% of individuals with PHP demonstrated a 238

successful response to 6 sessions of multi-modal physical therapy intervention 239

conducted over 4 weeks. If PHP symptoms were present for less than 7.2 240

months, the probability of a successful response increased to 68% and 83% for 241

the NPRS/FAAM/GRC and NPRS/FAAM criteria, respectively. There were no 242

other investigations found that examined success to intervention for PHP based 243

upon the 2 criteria used in this investigation. The criteria and rate of success to 244

conservative intervention for PHP varies widely between studies making direct 245

comparison difficult. Some investigations have reported success rates ranging 246

from 44% to 89% based upon patient reports of improvement of symptoms, no 247

pain, total/100% relief, resolution of symptoms, good to excellent results, or 80% 248

relief of symptoms.4, 14, 21, 39, 44, 47, 56, 59, 67 Other investigations have used outcome 249

instruments to define success that included an 11-point visual analog scale 250

(VAS) less than 5, GRC greater than +4, GRC greater than +5, or FAAM change 251

greater than 8 points and have reported success rates ranging from 45% to 252

80%.16, 35, 37 When looking at the success rates of individual measures from the 253

clinical trial (Table 1), rates of success were similar to other investigations that 254

demonstrated 80% success based on FAAM change greater than or equal to 8 255

and 70% success based on GRC greater than or equal to +5.16, 35 256

While no responder criteria was found to determine the cut point for 257

successful response in individuals with PHP, responder criteria based on multiple 258

measurement domains including pain, function, and patient’s global rating of 259

change has been developed and used in other populations.10, 49, 50 The 260

Page 10 of 25



For Peer Review

successful response criteria used in this investigation was chosen to minimize 261

the possibility of deeming a response successful based on only 1 domain of 262

response to treatment where unsuccessful responses may still be observed in 263

other domains. As a result, the success rates observed in this investigation may 264

be lower than investigations using only 1 measure to determine success. In 265

addition to using criteria involving 3 measures of response to treatment, a parallel 266

analysis was conducted only using the NPRS and FAAM criteria. While the 267

NPRS/FAAM/GRC and NPRS/FAAM criteria produced similar results in logistic 268

regression analysis, the NPRS/FAAM analysis resulted in an increased success 269

rate and a higher odds ratio associated with the predictor variable. The increased 270

success rate and odds ratio is likely attributed to the lower standards to achieve 271

success in the NPRS/FAAM criteria; ie, 2 versus 3 measures used as the cut 272

point for successful response. Despite evidence that the GRC lacked correlation 273

with functional change measures, GRC may represent an important treatment 274

response construct worthy of consideration when determining treatment 275

success.22, 57 In this study, a more conservative estimate of successful response 276

to treatment was obtained when the GRC was added to the pain and function 277

measures compared to the pain and function measures alone. Further 278

consensus is needed to improve consistency in reporting of response to PHP 279

conservative treatment so that better comparisons can be made between 280

investigations. 281

A successful response to physical therapy intervention was found in this 282

analysis despite a sample that included individuals who, on average, were obese 283

(Table 1).45 Furthermore, high BMI was not a predictor of poor response to 284

treatment in the logistic regression analysis. Previously, increased weight or 285

obesity had been associated with an increased risk of PHP.23, 27, 28, 54 In addition, 286

Page 11 of 25



For Peer Review

individuals who are overweight or obese were reportedly less responsive to 287

conservative treatment for PHP.21, 30, 67 Because PHP often limits weight-bearing 288

exercise, individuals with PHP are less able to use exercise to manage their 289

weight.48 The results of this analysis are promising in that individuals who are 290

overweight or obese can have a successful response to physical therapy 291

intervention for PHP allowing them to reap the benefits of weight-bearing 292

exercise. 293

Increased age is another factor that has been associated with PHP and 294

decreased response to conservative treatment.27, 51 In this analysis, age was not 295

a significant factor that predicted successful response to treatment, but this 296

sample was limited to individuals between the ages of 27 – 63. Because of the 297

limited age range of this sample, it is unclear if individuals greater than 63 years 298

of age have a different rate of success than individuals less than 63 years old. 299

Approximately 15% of individuals with PHP are greater 65 years or older and 300

therefore it is important to identify successful interventions for PHP in older 301

adults.55 302

The results of this analysis indicated that individuals with PHP symptom 303

duration less than 7 months were more likely to respond to the physical therapy 304

intervention included in the clinical trial. Two other studies were found that 305

reported a reduced response to conservative intervention for PHP if symptoms 306

were present for a longer duration; ie, 12 or more months.39, 67 In a telephone 307

survey of 100 patients treated conservatively by 3 orthopaedic surgeons, a good 308

result (defined as no symptoms) was achieved in 14/22 (64%) with symptoms for 309

12 or more months compared to 68/78 (87%) with symptoms less than 12 310

months (p < .003).67 Conservative treatment was case-specific and included 311

stretching, cushion or hard shoe inserts, anti-inflammatory medication or 312

Page 12 of 25



For Peer Review

injections, ice or heat, heel cup, night splint, or foot strapping. Similarly, in a 313

mailed follow-up survey to 157 PHP patients that received conservative 314

treatment from one orthopaedic surgeon, patients with symptoms greater than 12 315

months were less likely to have a good outcome defined as no pain (p < .04).39 316

Conservative treatment included anti-inflammatory medication or injection, night 317

splint, at least one session of exercise instruction provided by a physical 318

therapist, and a custom orthosis or heel cup. In the clinical trial that provided the 319

sample for this analysis, patients received physical therapy intervention that 320

included stretching and self mobilization exercises to the lower leg, ankle, and 321

foot in addition to manual therapy or electrophysiological modalities. Despite the 322

different conservative treatments provided to the participants that formed the 323

sample for this analysis and in other studies that demonstrated reduced 324

response with longer symptom duration, symptom duration appears to be a 325

significant prognostic factor to consider in the conservative management of 326

individuals with PHP. 327

Individuals with longer symptom duration may develop different pain 328

mechanisms that contribute to their PHP experience and response to treatment. 329

Individuals with a longer duration of pain demonstrated maladaptive psychosocial 330

factors consistent with a central sensitization versus nociceptive pain 331

mechanism.60 The reduced treatment response associated with longer duration 332

of PHP symptoms and evidence of lower foot function in PHP patients with 333

higher stress and depression indicates the potential for central sensitization 334

mechanisms in patients with a longer duration of PHP.12 Therefore, clinical 335

reasoning models that integrate intervention with consideration of pain 336

mechanism and clinical presentation may benefit individuals with longer PHP 337

symptoms. Therapeutic pain neuroscience education is one intervention that is 338

Page 13 of 25



For Peer Review

recommended in the presence of central sensitization and has demonstrated 339

positive effects on pain, disability, anxiety and stress in chronic musculoskeletal 340

pain conditions.36 No evidence was found on the effects of therapeutic pain 341

neuroscience education in individuals with PHP although a clinical trial is 342

underway that includes this as one part of the conservative management plan.40 343

In addition to therapeutic pain neuroscience education, other interventions 344

directed at proximal inputs to PHP through central or peripheral neurodynamic 345

mechanisms; eg, myofasicial trigger points, may be of additional benefit. 346

Myofascial trigger point manual and dry needling intervention has demonstrated 347

significant improvement of PHP, pain pressure thresholds, and foot-related 348

function.1, 11, 53 In investigations of myofascial trigger point therapy on PHP, only 349

Cotchett et al11 included a sample of participants with an average duration 350

greater than 7 months. Further research is needed to identify factors related to 351

longer duration of PHP and to test the effects of pain neuroscience education 352

and myofascial trigger point intervention on chronic PHP. 353

The results of this analysis indicated that symptom duration was a 354

significant predictor of response to physical therapy intervention for PHP. While, 355

the results of this analysis provide evidence of an improved response with a 356

shorter duration of symptoms and a lesser response with a longer duration of 357

symptoms, caution is warranted in generalizing the results. As indicated in Table 358

3, individuals with symptoms longer than 7 months had higher functional scores 359

on the FAAM and lower pain on the NPRS at baseline that may have limited 360

potential for improvement. Conversely, individuals with a lower duration of 361

symptoms had lower functional scores and higher pain scores at baseline which 362

may have contributed to greater potential for improvement. Therefore, because 363

the responder criteria were based on the magnitude of response; ie, the MCID of 364

Page 14 of 25



For Peer Review

outcome measures, the effect of duration determined from logistic regression 365

may have been affected by the baseline differences between individuals above 366

and below the 7 month duration cut point. Despite the statistical significance that 367

demonstrated greater odds of successful outcome in individuals with symptoms 368

less than 7 months, clinically meaningful changes were also observed in 369

individuals with symptoms greater than 7 months based on the MCID of the 370

NPRS and FAAM (Table 3). In addition, an equal proportion of individuals with 371

symptoms greater and less than 7 months demonstrated GRC of +5 or greater 372

(Table 3). While further research is warranted to examine effective solutions in 373

individuals with a longer duration of symptoms, the physical therapy intervention 374

provided in this clinical trial resulted in clinically-meaningful changes regardless 375

of symptom duration and was still effective in individuals with symptoms greater 376

than 7 months. 377

In addition to duration of symptoms, there may be other factors that were 378

not recorded during the clinical trial or considered in this analysis that may 379

predict successful treatment response. Symptom duration explained 11.7 to 380

32.6% of the variation in the data, and therefore other factors are likely to 381

contribute to the remaining variability in treatment response. Few studies were 382

found that investigated factors related to response to conservative treatment 383

other than age, BMI, and symptom duration. Several other factors derived from 384

the history and physical examination including foot posture/mobility, ankle or 385

hallux dorsiflexion, daily weight-bearing duration, lower leg/foot strength, 386

neurodynamic dysfunction, stress, depression, and low back pain have 387

demonstrated an association with PHP or foot function and may contribute to the 388

response to physical therapy intervention.3, 12, 26, 28, 34, 43, 54, 58, 64 Additional 389

information derived from imaging including heel pad energy dissipation ratio, 390

Page 15 of 25



For Peer Review

hyperemia, thickened plantar fascia, and heel spurs may also contribute to 391

prediction of response to treatment.27, 41, 63, 65, 66 While the historical, physical 392

examination and imaging factors listed above may occur more commonly in 393

some populations of PHP, they are also present in asymptomatic individuals and 394

the impact of these factors on clinical management including prediction of 395

successful response requires further investigation.17 396

397

398

Conclusion 399

400

Individuals with PHP for less than 7 months were more likely to respond to 401

the physical therapy interventions provided in the clinical trial, but age and BMI 402

were not predictive of treatment response. Despite the improved response of 403

individuals with symptoms less than 7 months, some individuals with symptoms 404

greater than 7 months were still able to demonstrate clinically meaningful 405

changes in response to physical therapy intervention. Participants in this trial 406

were obese based on the average BMI, yet were responsive to physical therapy 407

intervention despite previous reports of the associations of obesity with PHP and 408

a poor treatment response. The upper age range in this sample was limited and 409

further analysis is needed to assess the influence of age on treatment response 410

in individuals greater than 63 years old. Further research is needed to elucidate 411

additional strategies effective for PHP lasting longer than 7 months and matched 412

to individual patient characteristics and preferences. 413

414

415

References 416

417

Page 16 of 25



For Peer Review

1. Ajimsha MS, Binsu D, Chithra S. Effectiveness of myofascial release in the management 418

of plantar heel pain: A randomized controlled trial. Foot (Edinb). 2014;24:66-71. 419

2. Al Fischer Associates Inc. 2002 Podiatric Practice Survey. Statistical results. J Am Podiatr 420

Med Assoc. 2003;93:67-86. 421

3. Allen RH, Gross MT. Toe flexors strength and passive extension range of motion of the 422

first metatarsophalangeal joint in individuals with plantar fasciitis. J Orthop Sports Phys 423

Ther. 2003;33:468-478. 424

4. Amis J, Jennings L, Graham D, Graham CE. Painful heel syndrome: radiographic and 425

treatment assessment. Foot Ankle. 1988;9:91-95. 426

5. Beyzadeoglu T, Gokce A, Bekler H. The effectiveness of dorsiflexion night splint added to 427

conservative treatment for plantar fasciitis. Acta Orthop Traumatol Turc. 2007;41:220-428

224. 429

6. Buchbinder R. Clinical practice. Plantar fasciitis. N Engl J Med. 2004;350:2159-2166. 430

7. Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in patients 431

with low back pain. Spine (Phila Pa 1976). 2005;30:1331-1334. 432

8. Cleland JA, Abbott JH, Kidd MO, et al. Manual physical therapy and exercise versus 433

electrophysical agents and exercise in the management of plantar heel pain: a 434

multicenter randomized clinical trial. J Orthop Sports Phys Ther. 2009;39:573-585. 435

9. Cleland JA, Childs JD, Fritz JM, Whitman JM, Eberhart SL. Development of a clinical 436

prediction rule for guiding treatment of a subgroup of patients with neck pain: use of 437

thoracic spine manipulation, exercise, and patient education. Phys Ther. 2007;87:9-23. 438

10. Cleland JA, Fritz JM, Whitman JM, Heath R. Predictors of short-term outcome in people 439

with a clinical diagnosis of cervical radiculopathy. Phys Ther. 2007;87:1619-1632. 440

11. Cotchett MP, Munteanu SE, Landorf KB. Effectiveness of Trigger Point Dry Needling for 441

Plantar Heel Pain: A Randomized Controlled Trial. Phys Ther. 2014;94(8):1083-1094. 442

12. Cotchett MP, Whittaker G, Erbas B. Psychological variables associated with foot function 443

and foot pain in patients with plantar heel pain [published online March 20, 2014]. Clin 444

Rheumatol. doi 10.1007/s10067-014-2565-7. 445

13. Covey CJ, Mulder MD. Plantar fasciitis: How best to treat? J Fam Pract. 2013;62:466-471. 446

14. Davis PF, Severud E, Baxter DE. Painful heel syndrome: results of nonoperative 447

treatment. Foot Ankle Int. 1994;15:531-535. 448

15. Digiovanni BF, Nawoczenski DA, Malay DP, et al. Plantar fascia-specific stretching 449

exercise improves outcomes in patients with chronic plantar fasciitis. A prospective 450

clinical trial with two-year follow-up. J Bone Joint Surg Am. 2006;88:1775-1781. 451

16. Drake M, Bittenbender C, Boyles RE. The short-term effects of treating plantar fasciitis 452

with a temporary custom foot orthosis and stretching. J Orthop Sports Phys Ther. 453

2011;41:221-231. 454

17. Ehrmann C, Maier M, Mengiardi B, Pfirrmann CW, Sutter R. Calcaneal Attachment of the 455

Plantar Fascia: MR Findings in Asymptomatic Volunteers. Radiology. 2014;272:807-814. 456

18. Farrar JT, Young JP, Jr., LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes 457

in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain. 458

2001;94:149-158. 459

19. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis 460

program for the social, behavioral, and biomedical sciences. Behav Res Methods. 461

2007;39:175-191. 462

20. French HP, Galvin R, Cusack T, McCarthy GM. Predictors of short-term outcome to 463

exercise and manual therapy for people with hip osteoarthritis. Phys Ther. 2014;94:31-464

39. 465

21. Furey JG. Plantar fasciitis. The painful heel syndrome. J Bone Joint Surg Am. 466

1975;57:672-673. 467

22. Garrison C, Cook C. Clinimetrics corner: the Global Rating of Change Score (GRoC) poorly 468

correlates with functional measures and is not temporally stable. J Man Manip Ther. 469

2012;20:178-181. 470

Page 17 of 25



For Peer Review

23. Gill LH, Kiebzak GM. Outcome of nonsurgical treatment for plantar fasciitis. Foot Ankle 471

Int. 1996;17:527-532. 472

24. Hagen EM, Svensen E, Eriksen HR. Predictors and modifiers of treatment effect 473

influencing sick leave in subacute low back pain patients. Spine (Phila Pa 1976). 474

2005;30:2717-2723. 475

25. Harrell FE, Lee KL, Mark DB. Multivariable prognostic models: issues in developing 476

models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat 477

Med. 1996;15:361-387. 478

26. Hendrix CL, Jolly GP, Garbalosa JC, Blume P, DosRemedios E. Entrapment neuropathy: 479

the etiology of intractable chronic heel pain syndrome. J Foot Ankle Surg. 1998;37:273-480

279. 481

27. Irving DB, Cook JL, Menz HB. Factors associated with chronic plantar heel pain: a 482

systematic review. J Sci Med Sport. 2006;9:11-22; discussion 23-14. 483

28. Irving DB, Cook JL, Young MA, Menz HB. Obesity and pronated foot type may increase 484

the risk of chronic plantar heel pain: a matched case-control study. BMC Musculoskelet 485

Disord. 2007;8:41. 486

29. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the 487

minimal clinically important difference. Control Clin Trials. 1989;10:407-415. 488

30. Japour CJ, Vohra R, Vohra PK, Garfunkel L, Chin N. Management of heel pain syndrome 489

with acetic acid iontophoresis. J Am Podiatr Med Assoc. 1999;89:251-257. 490

31. Jensen MP, McFarland CA. Increasing the reliability and validity of pain intensity 491

measurement in chronic pain patients. Pain. 1993;55:195-203. 492

32. Jensen MP, Turner JA, Romano JM, Fisher LD. Comparative reliability and validity of 493

chronic pain intensity measures. Pain. 1999;83:157-162. 494

33. Jones ML, Evans N, Tefertiller C, et al. Activity-Based Therapy for Recovery of Walking in 495

Chronic Spinal Cord Injury: Results from a Secondary Analysis to Determine 496

Responsiveness to Therapy [published online August 4, 2014]. Arch Phys Med Rehabil. 497

doi: 10.1016/j.apmr.2014.07.401. 498

34. Kibler WB, Goldberg C, Chandler TJ. Functional biomechanical deficits in running 499

athletes with plantar fasciitis. Am J Sports Med. 1991;19:66-71. 500

35. Looney B, Srokose T, Fernandez-de-las-Penas C, Cleland JA. Graston instrument soft 501

tissue mobilization and home stretching for the management of plantar heel pain: a 502

case series. J Manipulative Physiol Ther. 2011;34:138-142. 503

36. Louw A, Diener I, Butler DS, Puentedura EJ. The effect of neuroscience education on 504

pain, disability, anxiety, and stress in chronic musculoskeletal pain. Arch Phys Med 505

Rehabil. 2011;92:2041-2056. 506

37. Lynch DM, Goforth WP, Martin JE, Odom RD, Preece CK, Kotter MW. Conservative 507

treatment of plantar fasciitis. A prospective study. J Am Podiatr Med Assoc. 508

1998;88:375-380. 509

38. Martin RL, Irrgang JJ, Burdett RG, Conti SF, Van Swearingen JM. Evidence of validity for 510

the Foot and Ankle Ability Measure (FAAM). Foot Ankle Int. 2005;26:968-983. 511

39. Martin RL, Irrgang JJ, Conti SF. Outcome study of subjects with insertional plantar 512

fasciitis. Foot Ankle Int. 1998;19:803-811. 513

40. McClinton SM, Flynn TW, Heiderscheit BC, et al. Comparison of usual podiatric care and 514

early physical therapy intervention for plantar heel pain: study protocol for a parallel-515

group randomized clinical trial. Trials. 2013;14:414. 516

41. McMillan AM, Landorf KB, Gregg JM, De Luca J, Cotchett MP, Menz HB. Hyperemia in 517

plantar fasciitis determined by power Doppler ultrasound. J Orthop Sports Phys Ther. 518

2013;43:875-880. 519

42. McPoil TG, Martin RL, Cornwall MW, Wukich DK, Irrgang JJ, Godges JJ. Heel pain--plantar 520

fasciitis: clinical practice guildelines linked to the international classification of function, 521

disability, and health from the orthopaedic section of the American Physical Therapy 522

Association. J Orthop Sports Phys Ther. 2008;38:A1-A18. 523

Page 18 of 25



For Peer Review

43. Menz HB, Dufour AB, Riskowski JL, Hillstrom HJ, Hannan MT. Foot posture, foot function 524

and low back pain: the Framingham Foot Study. Rheumatology (Oxford). 2013;52:2275-525

2282. 526

44. Mizel MS, Marymont JV, Trepman E. Treatment of plantar fasciitis with a night splint 527

and shoe modification consisting of a steel shank and anterior rocker bottom. Foot 528

Ankle Int. 1996;17:732-735. 529

45. United States Department of Health and Human Services. The practical guide: 530

Identification, evaluation, and treatment of overweight and obesity in adults. Bethesda, 531

MD: National institutes of Health publication 00-4084. 532

http://www.nhlbi.nih.gov/files/docs/guidelines/prctgd_c.pdf . Published October 2000. 533

Accessed September 28, 2014. 534

46. Nee RJ, Vicenzino B, Jull GA, Cleland JA, Coppieters MW. Baseline characteristics of 535

patients with nerve-related neck and arm pain predict the likely response to neural 536

tissue management. J Orthop Sports Phys Ther. 2013;43:379-391. 537

47. O'Brien D, Martin WJ. A retrospective analysis of heel pain. J Am Podiatr Med Assoc. 538

1985;75:416-418. 539

48. Orchard J. Plantar fasciitis. BMJ. 2012;345:e6603. 540

49. Pham T, van der Heijde D, Altman RD, et al. OMERACT-OARSI initiative: Osteoarthritis 541

Research Society International set of responder criteria for osteoarthritis clinical trials 542

revisited. Osteoarthritis Cartilage. 2004;12:389-399. 543

50. Pham T, Van Der Heijde D, Lassere M, et al. Outcome variables for osteoarthritis clinical 544

trials: The OMERACT-OARSI set of responder criteria. J Rheumatol. 2003;30:1648-1654. 545

51. Probe RA, Baca M, Adams R, Preece C. Night splint treatment for plantar fasciitis. A 546

prospective randomized study. Clin Orthop Relat Res. 1999;190-195. 547

52. Reischl S. Physical therapist foot care survey. Orthop Pract. 2001;13:27. 548

53. Renan-Ordine R, Alburquerque-Sendin F, de Souza DP, Cleland JA, Fernandez-de-Las-549

Penas C. Effectiveness of myofascial trigger point manual therapy combined with a self-550

stretching protocol for the management of plantar heel pain: a randomized controlled 551

trial. J Orthop Sports Phys Ther. 2011;41:43-50. 552

54. Riddle DL, Pulisic M, Pidcoe P, Johnson RE. Risk factors for Plantar fasciitis: a matched 553

case-control study. J Bone Joint Surg Am. 2003;85-A:872-877. 554

55. Riddle DL, Schappert SM. Volume of ambulatory care visits and patterns of care for 555

patients diagnosed with plantar fasciitis: a national study of medical doctors. Foot Ankle 556

Int. 2004;25:303-310. 557

56. Scherer PR. Heel spur syndrome. Pathomechanics and nonsurgical treatment. 558

Biomechanics Graduate Research Group for 1988. J Am Podiatr Med Assoc. 1991;81:68-559

72. 560

57. Schmitt JS, Abbott JH. Patient Global Ratings of Change Did Not Adequately Reflect 561

Change Over Time: A Clinical Cohort Study. Phys Ther. 2014;94(4):534-542. 562

58. Schon LC, Glennon TP, Baxter DE. Heel pain syndrome: electrodiagnostic support for 563

nerve entrapment. Foot Ankle. 1993;14:129-135. 564

59. Shikoff MD, Figura MA, Postar SE. A retrospective study of 195 patients with heel pain. J 565

Am Podiatr Med Assoc. 1986;76:71-75. 566

60. Smart KM, Blake C, Staines A, Thacker M, Doody C. Mechanisms-based classifications of 567

musculoskeletal pain: part 1 of 3: symptoms and signs of central sensitisation in patients 568

with low back (± leg) pain. Man Ther. 2012;17:336-344. 569

61. Thomas JL, Christensen JC, Kravitz SR, et al. The diagnosis and treatment of heel pain: a 570

clinical practice guideline-revision 2010. J Foot Ankle Surg. 2010;49:S1-19. 571

62. Tong KB, Furia J. Economic burden of plantar fasciitis treatment in the United States. Am 572

J Orthop (Belle Mead NJ). 2010;39:227-231. 573

63. Wearing SC, Smeathers JE, Sullivan PM, Yates B, Urry SR, Dubois P. Plantar fasciitis: are 574

pain and fascial thickness associated with arch shape and loading? Phys Ther. 575

2007;87:1002-1008. 576

Page 19 of 25



For Peer Review

64. Wearing SC, Smeathers JE, Urry SR, Hennig EM, Hills AP. The pathomechanics of plantar 577

fasciitis. Sports Med. 2006;36:585-611. 578

65. Wearing SC, Smeathers JE, Urry SR, Sullivan PM, Yates B, Dubois P. Plantar 579

enthesopathy: thickening of the enthesis is correlated with energy dissipation of the 580

plantar fat pad during walking. Am J Sports Med. 2010;38:2522-2527. 581

66. Wearing SC, Smeathers JE, Yates B, Urry SR, Dubois P. Bulk compressive properties of 582

the heel fat pad during walking: a pilot investigation in plantar heel pain. Clin Biomech 583

(Bristol, Avon). 2009;24:397-402. 584

67. Wolgin M, Cook C, Graham C, Mauldin D. Conservative treatment of plantar heel pain: 585

long-term follow-up. Foot Ankle Int. 1994;15:97-102. 586

68. Yi TI, Lee GE, Seo IS, Huh WS, Yoon TH, Kim BR. Clinical characteristics of the causes of 587

plantar heel pain. Ann Rehabil Med. 2011;35:507-513. 588

589

Page 20 of 25



For Peer Review

Legends 590

591

Table 1: Participant characteristics and outcome measures at baseline and the 6 month follow-up (mean ± SD or frequency (%); n=60).

Variable Baseline

6 month Difference (95% CI)

Age in years (range) 48 ± 9 (27 - 63) - - Female participants (%) 33 (55) - - BMI in kg/m2 31.9 ± 7.5 - - Duration of symptoms in days 271.6 ± 212.7 - - NPRS 4.7 ± 1.7 1.7 ± 1.82 -3 (-2.4, -3.6) Change ≥ 2 points (%) - 42 (70) -

FAAM 57.5 ± 14.9 80 ± 16.5 22.5 (16.8, 28.2) Change ≥ 8 points (%) - 46 (77) - GRC ≥ +5 (%) - 40 (67) - Successful response (%) NPRS/FAAM/GRC criteria - 29 (48) - NPRS/FAAM criteria - 38 (63) - Abbreviations: BMI, body mass index; FAAM, Foot and Ankle Ability Measure; GRC, Global Rating of Change; NPRS, Numeric Pain Rating Scale.

592

593

Table 2: Odds ratio, sensitivity, specificity, and positive and negative likelihood ratio values of symptom duration less than 7.2 months as a predictor of successful response to physical therapy intervention.

Successful response criteria

Odds Ratio

(95% CI)

Sensitivity (95% CI)

Specificity (95% CI)

+ LR (95% CI)

- LR (95% CI)

NPRS/FAAM/GRC 4.2 (1.3, 13.8)

0.58 (0.41, 0.74)

0.75 (0.57, 0.87)

2.3 (1.2, 4.7)

0.56 (0.35, 0.89)

NPRS/FAAM 8.5 (2.5, 28.9)

0.76 (0.6, 0.87)

0.73 (0.52, 0.87)

2.8 (1.4, 5.6)

0.33 (0.18, 0.62)

Abbreviations: FAAM, Foot and Ankle Ability Measure; GRC, Global Rating of Change; LR, likelihood ratio; NPRS, Numeric Pain Rating Scale.

594

595

Page 21 of 25



For Peer Review

596

Table 3: Outcome comparison of individuals with symptom duration above and below the 7.2 month cut point (mean ± SD or frequency (%); n=60). The p-values indicate the lowest value from MANCOVA for all imputed data sets unless otherwise indicated.

Outcome Variable Symptoms < 7.2 months (n=35)

Symptoms > 7.2 months (n=25)

Difference (95% CI)

p-value

FAAM Baseline 53.3 ± 15.4 63.2 ± 15 -9.9 (-1.9, -18) <.001 6 month 82 ± 17.2* 77.2 ± 17.5* 4.8 (-4.3, 13.9)* > .19* Change score 24.5 ± 17.2* 19.6 ± 17.5* 4.4 (-4.7, 13.5)* > .19* NPRS Baseline 5.5 ± 1.6 3.7 ± 1.5 1.8 (1, 2.6) <.001 6 month 1.4 ± 2.2* 2.1 ± 2.1* -0.7 (-1.8, 0.4)* > .08* Change score -4 ± 2.2* -1.7 ± 2.1* -2.3 (-1.2, -3.4)* > .08* GRC (%) 6 month 25 (68.6) 16 (64) - > .36† Abbreviations: FAAM, Foot and Ankle Ability Measure; GRC, Global Rating of Change; NPRS, Numeric Pain Rating Scale. *Adjusted for baseline score. †Calculated via Chi-square test. 597

Page 22 of 25



For Peer Review

Paper No. FAI-14-0438.R1 entitled "Predictors of Response to Physical Therapy

Intervention for Plantar Heel Pain: A Secondary Analysis of Data from a Randomized

Clinical Trial."

Reviewer #1:

I'm still not satisfied with what disease state is being treated in this study. Plantar heel pain

is not always plantar fascitis. No imaging was attempted to discern a diagnosis outside of

"plantar heel pain".

RESPONSE TEXT CHANGES

In response to this comment and suggestions of the Associate

Editor, the inclusion/exclusion criteria have been more explicitly

defined in the Materials and Methods section.

Lines 84 - 88

In addition, further details have been added to the Introduction

to provide upfront clarification that plantar fasciitis is one of

many potential reasons for plantar heel pain, but is not the only

condition that produces plantar heel pain.

Lines 44 - 47

Additional limitations were added at the end of the discussion

to include consideration of factors that would be identified by

imaging.

Lines 389 - 396

COMMENT RESPONSE TEXT CHANGES

Abstract

I'm still not satisfied with what

disease state is being treated in this

study. Plantar heel pain is not

always plantar fascitis. No imaging

was attempted to discern a

diagnosis outside of "plantar heel

pain".

Please see response

above.

Please see text changes

indicated above.

Introduction







pain".

Please see response

above.


indicated above.

Methods







pain".

Please see response

above.


indicated above.

Results

Page 23 of 25



For Peer Review







pain".

Please see response

above.


indicated above.

Discussion







pain".

Please see response

above.


indicated above.

Conclusion







pain".

Please see response

above.


indicated above.

Associate Editor: Campbell, John

Comments to the Author:

The authors should be congratulated on an outcome study that investigates a very common

condition, plantar heel pain. I still share Reviewer #1's concerns which seem valid. The

authors do not provide any inclusion/diagnostic criteria for PHP (our readers are mainly

clinicians who treat numerous pathologies...PF? tarsal tunnel? Baxter's nerve extrapment?

fat pad atrophy?). I think the diagnostic criteria must be more explicitly defined. (this is what

Reviewer #1 was getting at in asking about imaging)

Thank you for the additional clarification of Reviewer #1’s concerns. Please see the response

to reviewer #1’s comment and text changes indicated above.

Were there any other treatment methods? NSAIDs? Night splint? Injections? Immobilization

(boot or cast)? You must clarify this or it will be seen as an unspoken confounding factor.

No other interventions were provided other than those listed in the Materials and Methods

section. Additional information has been added to indicate the number of participants who

were taking medication at the start of the trial and to indicate that no other interventions

were utilized. Please see lines 138 – 145.

Additionally, I still have trouble accepting the Methods, which treated the pateints with two

different regimens but then combined the data to look at correlation with BMI, age, etc.

Although there may be some statistical precedence for this, I believe many of our readers

Page 24 of 25



For Peer Review

will look skeptically on this. This would likely need to be more clearly explained for my

satisfaction.

Additional information has been added to the Statistical Analysis section to include

references to recommended procedures for this type of analysis and other studies that have

used similar procedures. Please see lines 150 – 159.

Page 25 of 25



predictors of response to physical therapy intervention for · predictors of response to physical...

Documents