Prateek Sharma1, Rosemin Kassam1, Jason Sutherland1, Malcolm Maclure1, David Patrick1,3Evaluation of a population­level academic detailing intervention on UTIs in British Columbia’s nursing homes

1School of Population and Public Health, University of British Columbia 2Department of Anesthesiology, Pharmacology and Therapeutics, University of British Columbia 3British Columbia Centre for Disease Control

• In 2014, the volume of prescribing in British Columbia's (BC) nursing homes was double the general populations' (35 vs 16 DDD per 1000 population per day)1,2

• 18.5% of residents prescribed antibiotics were treated for urinary tract infections (UTI); the most commonly linked indication for an antibiotic in 20142

• However, up to 55% of antibiotics prescribed for UTIs in nursing homes can be considered inappropriate3

• Knowledge translation efforts from our earlier research motivated the BC Provincial Academic Detailing (PAD) to address unnecessary treatment of UTIs in BC's 301 nursing home facilities4,5

• Is PAD’s academic detailing intervention associated with a decrease in the trend of antibiotics linked to a UTI indication?

• We hypothesized that the intervention group's trend decreased in the post­intervention period, while the control's groups pre­existing trends would continue as before.

Research question and hypothesis

References

Academic detailing intervention• From June to December, 2016, eight pharmacist­trained academic detailers from

BC PAD visited 131 publicly­subsidized nursing homes.• Physicians, nurses, nurse practitioners and care aides attended metings.• Key messages were:

1 Do not assume that non­specific symptoms are a UTI.2 Dipstick urinalysis and urine cultures should not be ordered unless there are

urinary tract infection­specific symptoms present.3 Most people with asymptomatic bacteriuria should not be treated with

antibiotics.4 Nitrofurantoin is preferred empirically for acute cystitis (bladder infection) in

women but not pyelonephritis or more complicated UTIs.

Urinary tract infections in nursing homes

Interrupted time series evaluation

Figure 2. Data source and evaluation outcome

Administrative database of prescriptions dispensed in community pharmacies

Data on services provided by fee­for­service practitioners

Outcome:

1. BCCDC. Antimicrobial Utilization. 2017, http://www.bccdc.ca/health­professionals/data­reports/antimicrobial­surveillance­tools/antimicrobial­utilization­dashboard/antimicrobial­utilization­executive­summary. 2. Marra, Fawziah, et al. "Utilization of Antibiotics in Long­Term Care Facilities in British Columbia, Canada." J. Am. Med. Dir. Assoc. 18.12 (2017): 1098­e1. 3. Eure, Taniece, et al. "Measuring Antibiotic Appropriateness for Urinary Tract Infections in Nursing Home Residents." Infect. Control Hosp. Epidemiol. 38.8 (2017): 998­1001. 4. Canadian Institute for Health Information. Profile of Residents in Residential and Hospital­Based Continuing Care 2015­2016. 2016, https://www.cihi.ca/en/profile­of­residents­in­residential­and­hospital­based­continuing­care­2015­2016. 5. Provincial Academic Detailing Service. UTIs in Primary and Long Term Care. 2017, http://www2.gov.bc.ca/gov/content/health/practitioner­professional­resources/pad­service/utis­in­primary­and­long­term­care. 6. Karaca­Mandic, Pinar, Edward C. Norton, and Bryan Dowd. "Interaction terms in nonlinear models." Health Serv. Res. 47.1pt1 (2012): 255­274. 7. Loeb, Mark, et al. "Effect of a multifaceted intervention on number of antimicrobial prescriptions for suspected urinary tract infections in residents of nursing homes: cluster randomised controlled trial." Bmj 331.7518 (2005): 669. 8. Morrill, Haley J., et al. "Antimicrobial stewardship in long­term care facilities: a call to action." J. Am. Med. Dir. Assoc. 17.2 (2016): 183­e1.

• Inclusion criteria: antibiotics prescribed to residents covered through Pharmacare Plan B (nursing home insurance scheme) from June 2015 to March 2017 with UTI­related International Classification of Diseases, Ninth Revision codes in the Medical Services Plan

• Our outcome measure was days of supply (DOS). Monthly data was analyzed with a multilevel negative­binomial segmented regression with random effects for LHA, pre­ and post­intervention trends. To better interpret results of the non­linear model, we assessed interaction on an additive scale6

• The University of British Columbia Ethics Review Board approved the project

Figure 1. Coverage of nursing homes in BC's local health areas (LHAs) by BC PAD from June to December, 2016

• Presently, this is one of the few large­scale studies of antibiotic stewardship in nursing homes7. This paucity of context­specific evidence purportedly hinders more widespread implementation of antibiotic stewardship in nursing homes8.

• We are awaiting denominator data to improve the multilevel model and better interpret the trends.

• Our analysis looked only at prescribing by fee­for­service providers, which may bias results from the general population. However, a sensitivity analysis for linkage bias indicated that results were robust to this bias (data not shown).

Results (continued)

Table 2. Characteristics of nursing home resident and Pharmanet prescribing cohorts, June 2015 to March 2017

Results

Results (continued)

• During the study period, 8,686 residents received 25,141 UTI­linked prescriptions.

• While the expected post­intervention rate for the intervention group was ­1.1 [­1.8, ­0.3] DOS per month, the observed trend was ­2.8 [­2.8, ­0.7] DOS per month; 169.9% lower than expected [­59.7%, 663.7%]. The control’s average post­intervention trend was unchanged, ­0.1 [­0.6, 0.2] DOS per month.

Results summary

Table 1. Characteristics of prescribed antibiotics in Pharmanet Plan B, June 2015 to March 2017

Figure 3. Average model fitted days of supply of UTI­linked prescriptions in intervention and control LHAs, June 2015 to March 2017

ConclusionIn this pragmatic ecologic evaluation, academic detailing was associated with reductions in UTI­coded antibiotic prescribing.

Strengths and limitations:

*Proportion of UTI­linked prescriptions

Table 3. Additive monthly rates of change of UTI­linked prescribing averaged by intervention status and period

Discussion

No nursing home in LHA