Antibiotic use in long-term care facilities

Transcription

1 J Antimicrob Chemother 2011; 66: doi: /jac/dkr395 Advance Access publication 27 September 2011 Antibiotic use in long-term care facilities Nick Daneman 1,2 *, Andrea Gruneir 1,3,4, Alice Newman 1, Hadas D. Fischer 1, Susan E. Bronskill 1,4, Paula A. Rochon 1,3,4, Geoff M. Anderson 1,3,4 and Chaim M. Bell 1,5 1 Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada; 2 Division of Infectious Diseases, Department of Medicine, Sunnybrook Health Sciences Centre, University of Toronto, Ontario, Canada; 3 Women s College Research Institute, Women s College Hospital, University of Toronto, Ontario, Canada; 4 Department of Health Policy, Management and Evaluation, University of Toronto, Ontario, Canada; 5 Keenan Research Centre in the Li Ka Shing Knowledge Institute and Division of General Internal Medicine, St Michael s Hospital, University of Toronto, Ontario, Canada *Corresponding author. Division of Infectious Diseases, Sunnybrook Health Sciences Centre, University of Toronto, 2075 Bayview Ave, Toronto, Ontario, Canada M4N 2M5. Tel: x2791; Fax: ; nick.daneman@sunnybrook.ca Received 24 May 2011; returned 11 July 2011; revised 12 July 2011; accepted 30 August 2011 Objectives: Evaluation and optimization of antibiotic use (antibiotic stewardship) is being increasingly promoted as a means to reduce antibiotic resistance, adverse events, treatment complications and costs within institutions. Our goal was to examine the prevalence of antibiotic use among long-term care facility residents and the extent of variability across these institutions. Methods: We conducted a population-based, point-prevalence study of antibiotic use among elderly individuals (n¼37371) residing in long-term care facilities (n¼363 institutions) in Ontario between April and June 2009, using linked healthcare databases from Canada s largest province. Facilities were grouped into quintiles according to their mean antibiotic dispensing rates and variation was compared across quintiles. Results: There were 2190 (5.9%) long-term care residents receiving antibiotic prescriptions on the study date. The three most prevalent antibiotics were agents most commonly used for the treatment of urinary tract infections, including nitrofurantoin (365, 15.4%), trimethoprim/sulfamethoxazole (338, 14.3%) and ciprofloxacin (304, 12.8%). The majority of treatment courses were at least 10 days in duration (1482, 62.6%), and many exceeded 90 days (495, 20.9%), suggesting chronic prophylaxis. There was substantial variability in antibiotic use across facilities, with a 5-fold variation from the highest-use quintile (10.8%) to the lowest-use quintile (2.2%). This variation persisted after adjustment for multiple facility-level and resident-level factors, including demographic characteristics, healthcare utilization statistics, co-morbidity prevalence, functional status and device dependence. Conclusions: Antibiotic use is common among long-term care residents, variable across institutions, and may benefit from focused antimicrobial stewardship interventions to standardize treatment indications and duration. Keywords: antibacterials, nursing homes, health services research, quality indicators Introduction It is estimated that up to half of antibiotic use in hospitals is unnecessary or inappropriate. 1,2 As with other pharmaceutical classes, inappropriate use is associated with avoidable adverse events for the recipient; unlike other pharmaceutical classes, overuse of antibiotics also jeopardizes other hospitalized patients (present and future) through selection of antibiotic-resistant pathogens. 1,3 Relentless increases in antibiotic utilization in hospitals 4,5 are driving antibiotic resistance among hospital-acquired pathogens. 6,7 At the same time, the production of new pharmaceutical agents is waning, leaving clinicians with limited (and sometimes no) effective antibiotic options for their patients. 3,8 Infections with antibiotic-resistant organisms have already been associated with increased attributable length of stay, mortality and costs. 9 As a result, the Infectious Diseases Society of America (IDSA) and the Society for Healthcare Epidemiology of America (SHEA) have recently published joint recommendations and guidelines for the development of institutional antimicrobial stewardship programmes. 1 Antimicrobial stewardship has been defined as the limitation of inappropriate antimicrobial use while optimizing antimicrobial drug selection, dosing, route and duration of therapy in order to maximize clinical cure and to limit unintended consequences, such as the emergence of resistance, # The Author Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please journals.permissions@oup.com 2856

2 Antibiotic use in nursing homes JAC adverse drug events and the selection of pathogenic organisms (such as Clostridium difficile). 1 Antibiotic stewardship may be just as pressing in long-term care facilities (also known as nursing homes). These facilities are also home to vulnerable, elderly and crowded residents, and antibiotics are among the most commonly prescribed classes of medications for these residents. Just as in the hospital setting, it is estimated that about half of long-term care facility antibiotic prescriptions are unnecessary or inappropriate. 10 Moreover, long-term care residents are in frequent contact with acute care hospitals, 11 so the current attempts to rationalize antibiotic use within hospitals will be hampered without similar evaluations and interventions to optimize antibiotics in these longterm care facilities. 10 The objective of this population-based study was to describe the prevalence and variation in antibiotic use across long-term care facilities and identify potential targets for future antibiotic stewardship interventions. Materials and methods Study design and data sources We conducted a point-prevalence study of antibiotic use among elderly residents of long-term care facilities in Ontario, Canada, via linked administrative databases at the Institute for Clinical Evaluative Sciences (ICES). These well-validated databases have been used extensively in prior research, including multiple investigations of medication use among elderly individuals. 12,13 These datasets are linked via encrypted healthcare numbers and include: (i) the Continuing Care Reporting System Long Term Care (CCRS-LTC) database, which includes detailed quarterly assessments for all Ontario long-term care residents; (ii) the Ontario Drug Benefit Program (ODB) database, which contains detailed drug information for Ontario s.1 million elderly; (iii) the Registered Persons database (RPDB), which contains demographic data for all of Ontario s 12.5 million residents; (iv) the Ontario Health Insurance Plan (OHIP) database, which includes physician billing claims for visits and procedures performed within Ontario s universal single-payer healthcare system; (v) the Canadian Institute for Health Information Discharge abstract database (CIHI-DAD), which details all hospitalization events in the province; and (vi) the National Ambulatory Care Reporting System (NACRS) database, which describes all emergency room visits. Resident selection criteria We identified all elderly Ontarians ( 66 years old) residing in a long-term care facility who had a completed CCRS-LTC assessment between 1 April and 30 June The CCRS-LTC assessments are performed using the Resident Assessment Instrument (RAI) Minimum Data Set (MDS 2.0), designed to provide a comprehensive clinical assessment of the functional status and care needs of long-term care residents. 14 This tool exhibits excellent reliability and internal consistency. 14 Given that assessments are completed at minimum quarterly intervals, we expect that every resident would have undergone at least one assessment during this 3 month period. Only the first MDS assessment was included for each individual resident within the study period. Antibiotic utilization Antibiotic prescriptions were ascertained from the ODB database, which includes information on drug name, class, dose, route, interval and duration. This database is both comprehensive (ODB provides universal medication coverage for all Ontario seniors 65 years old) and accurate (concordance with pharmacy chart review exceeds 99%). 15 We included all systemic antibiotics administered via the enteral or parenteral route and excluded local and topical antibiotic treatments. Antibiotic prescriptions were grouped by generic drug names, as well as into the following pharmacological classes (and subclasses): penicillins, cephalosporins (first generation, second/third generation), fluoroquinolones (second generation, third generation), macrolides, trimethoprim and/or sulphonamides, tetracyclines, lincosamides, nitrofurantoin, metronidazole, aminoglycosides, glycopeptides, oxazolidinones and carbapenems. The point prevalence of antibiotic use was calculated by estimating the proportion of all residents with an index MDS assessment who also had an antibiotic prescription overlapping the date of that index assessment. For example, if a resident s index assessment date was 16 May 2009 and s/he had received a prescription for a 5 day course of antibiotics on 14 May 2009, then that resident was counted in the numerator of the point-prevalence estimate. If residents received more than one antibiotic, both of these were included in the descriptive analysis of antibiotic utilization, but were counted only once in the numerator of antibiotic point-prevalence estimates. The duration of each individual antibiotic prescription was also measured from ODB, and if the same drug was renewed within 3 days of completion, then all consecutive treatment courses were summed to determine the total duration of therapy. Antibiotic durations of 90 days were collapsed into a single category as being indicative of chronic therapy or prophylaxis. As an additional descriptor, antibiotic duration of treatment was dichotomized into short-course (,10 days) and long-course therapy ( 10 days). Facility- and resident-level risk factors The long-term care facility home of each individual resident was identified from the CCRS-LTC, including facility size (number of residents) and location (urban versus rural location). An extensive number of residentlevel factors were collected from the index assessment and a 1 year look back in provincial drug, hospital and emergency room databases. Demographic variables included age, sex and time since admission to the facility. Healthcare utilization variables included number of prescription medications, emergency department visits and inpatient hospital admissions over the preceding year, as well as documentation of do not resuscitate or do not hospitalize orders. Co-morbidities of interest included diabetes mellitus, atherosclerotic heart disease, congestive heart failure, peripheral vascular disease, dementia, stroke, paralysis, Parkinson s disease, obstructive lung disease, cancer, gastrointestinal disease, liver disease and renal disease. Measures of function included the proportion of residents with urinary or faecal incontinence and the proportion needing assistance to accomplish each of the five activities of daily living. Use of a number of procedures and devices was identified, including urinary catheters, dialysis, intravenous medications, tracheostomies, ventilators and feeding tubes. Statistical analysis Descriptive analyses were performed to examine the characteristics of Ontario long-term care facilities and their residents, as well as the prevalence, types and duration of antibiotics prescribed to these residents. After excluding facilities with,20 residents, the remaining facilities were classified into quintiles based on the point prevalence of antibiotic prescriptions among their residents during the study period: quintile 1 represented the 20% of facilities with the lowest antibiotic prevalence, whereas quintile 5 represented the 20% of facilities with the highest antibiotic prevalence. Prevalence ratios were calculated to compare antibiotic prescribing rates in the higher-use quintiles to rates in quintile 1. Logistic regression was used to assess the association between the facility antibiotic prevalence quintile and an individual resident s likelihood of being prescribed an antibiotic on the index assessment date. The facility antibiotic quintile was entered into the model as a categorical variable, 2857

3 Daneman et al. with the lowest quintile treated as the reference category. Other resident and facility characteristics were included in the model to assess whether the effect of the facility antibiotic prevalence quintile persisted after controlling for these factors. The Cochran Armitage trend test was used to evaluate potential increases in use of long-duration ( 10 day) antibiotic therapy across antibiotic prevalence quintiles. All analyses were performed using SAS statistical software version 9.2 (SAS, Cary, NC, USA). Patient confidentiality was maintained via encrypted health card numbers using ICES protocols. The study was approved by the research ethics board of Sunnybrook Health Sciences Centre. Table 1. Characteristics of the Ontario long-term care facility residents Characteristic Number of residents (%) Demographics median age, years (IQR) 85 (80 90) female (72.2) recent admission within 90 days 1540 (4.1) median time since admission, years (IQR) 1.7 ( ) Health system use prescription drugs in past 12 months a 12 (8 17) emergency room visits in past 12 months (30.2) inpatient admissions in past 12 months 9618 (25.7) do not resuscitate order (76.6) do not hospitalize order 6862 (18.4) Co-morbidities diabetes mellitus 8792 (23.5) atherosclerotic heart disease 4613 (12.3) congestive heart failure 4272 (11.4) peripheral vascular disease 1828 (4.9) dementia (59.5) stroke 7737 (20.7) paralysis 1964 (5.3) Parkinson s disease 2658 (7.1) obstructive lung disease 5774 (15.5) cancer 3288 (8.8) gastrointestinal disease 5541 (14.8) liver disease 264 (0.7) renal disease 3061 (8.2) Functional status requires assistance for transferring (67.3) requires assistance for dressing (82.8) requires assistance for eating (35.4) requires assistance for toileting (75.5) requires assistance for hygiene (82.0) bowel incontinence (58.6) bladder incontinence (77.2) Devices urinary catheter 1037 (2.8) dialysis 215 (0.6) intravenous medications 156 (0.4) tracheostomy 91 (0.2) respiratory ventilator 26 (0.1) feeding tube 369 (1.0) a Median (IQR). Results Description of long-term care facility residents in Ontario The final cohort included prevalent elderly long-term care facility residents. These residents were distributed across 363 long-term care facilities, located within predominantly urban settings (80.2%), and housing an average of 103 residents per facility (interquartile range ). The long-term care facility residents were mostly women (72%), with a median age of 85 years, median 1.7 years stay in long-term care, and nearly three-fifths (59%) having a dementia diagnosis (Table 1). Most residents were incontinent and required assistance with activities of daily living, but only a minority was reported to have indwelling catheters or other medical devices (Table 1). Prevalence of antibiotic use among long-term care residents Among this cohort of long-term care residents, 2190 (5.9%) were receiving antibiotics at the time of their index assessment, and since some residents were receiving more than one antibiotic, the total number of distinct prescriptions was 2366 (Table 2). Only 961 (43.9%) of the antibiotic recipients had an associated claim for a physician bedside visit within 1 day of their index prescription, and only 370 (16.9%) had a physician claim with a Table 2. Most frequently prescribed antibiotics among Ontario long-term care facility residents Antibiotic name Number of antibiotic treatment courses (%) Nitrofurantoin 365 (15.4) Trimethoprim/sulfamethoxazole 338 (14.3) Ciprofloxacin 304 (12.8) Cefalexin 241 (10.2) Amoxicillin 235 (9.9) Moxifloxacin 134 (5.7) Norfloxacin 123 (5.2) Levofloxacin 105 (4.4) Clarithromycin 83 (3.5) Cefuroxime 73 (3.1) Metronidazole 73 (3.1) Cloxacillin 59 (2.5) Tetracycline 56 (2.4) Azithromycin 47 (2.0) Clindamycin 36 (1.5) Cefprozil 22 (0.9) Penicillin 17 (0.7) Cefixime 16 (0.7) Ceftriaxone 12 (0.5) Ampicillin 10 (0.4) Erythromycin 10 (0.4) Vancomycin 5 (0.2) Cefaclor 5 (0.1) Total 2366 (100) 2858

4 Nitrofurantoin Sulphonamides Second-generation fluoroquinolones Penicillins Metronidazole Macrolides Lincosamides Antibiotic use in nursing homes Number of antibiotic prescriptions Third-generation fluoroquinolones Tetracyclines Second/third-generation cephalosporin First-generation cephalosporin Total duration of treatment (days) Figure 1. Distribution of antibiotic prescription durations. The majority of antibiotic prescriptions are prescribed in multiples of 5 or 7 days, with the most common durations being 7, 10 and 14 days. Large numbers of antibiotic prescriptions are chronic ( 90 days in duration). Colour segments represent different antibiotic classes and subclasses JAC

5 Daneman et al. Table 3. Facility- and resident-level characteristics across facility quintiles based on point prevalence of antibiotic use Characteristic Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Facility characteristics number of facilities total number of residents mean residents per facility (SD) (69.9) (83.1) (104.8) (50.3) (44.5) % of residents on antibiotics, mean (SD) 2.2 (0.9) 4.2 (0.4) 5.7 (0.4) 7.4 (0.6) 10.8 (2.6) Resident characteristics a age, years (median, IQR) 85 (80 90) 85 (80 90) 86 (80 90) 86 (81 90) 85 (80 90) female 4421 (70.4) 6567 (73.8) 6464 (70.8) 5202 (72.7) 4089 (73.0) recent admission within 90 days 236 (3.8) 339 (3.8) 366 (4.0) 301 (4.2) 278 (5.0) years admitted (median, IQR) 2 (1 4) 2 (1 4) 2 (1 4) 2 (1 4) 2 (0 4) Health system use prescription drugs in past year b 12 (8 16) 12 (8 16) 12 (8 17) 13 (9 17) 13 (9 17) emergency room visits 1835 (29.2) 2605 (29.3) 2635 (28.9) 2281 (31.9) 1816 (32.4) inpatient admissions 1654 (26.3) 2175 (24.4) 2280 (25.0) 1876 (26.2) 1543 (27.6) do not resuscitate order 4683 (74.5) 6793 (76.4) 7141 (78.3) 5389 (75.3) 4379 (78.2) do not hospitalize order 1001 (15.9) 1482 (16.7) 1703 (18.7) 1406 (19.6) 1200 (21.4) Co-morbidities diabetes mellitus 1619 (25.8) 2118 (23.8) 2039 (22.3) 1651 (23.1) 1301 (23.2) atherosclerotic heart disease 943 (15.0) 1101 (12.4) 1065 (11.7) 904 (12.6) 569 (10.2) congestive heart failure 705 (11.2) 933 (10.5) 1046 (11.5) 915 (12.8) 640 (11.4) peripheral vascular disease 291 (4.6) 494 (5.6) 450 (4.9) 361 (5.0) 224 (4.0) dementia 3628 (57.7) 5226 (58.7) 5732 (62.8) 4340 (60.6) 3147 (56.2) stroke 1309 (20.8) 1764 (19.8) 2001 (21.9) 1469 (20.5) 1130 (20.2) paralysis 341 (5.4) 453 (5.1) 556 (6.1) 299 (4.2) 300 (5.4) Parkinson s disease 466 (7.4) 576 (6.5) 720 (7.9) 495 (6.9) 375 (6.7) obstructive lung disease 979 (15.6) 1352 (15.2) 1449 (15.9) 1154 (16.1) 794 (14.2) cancer 511 (8.1) 773 (8.7) 833 (9.1) 631 (8.8) 506 (9.0) gastrointestinal disease 876 (13.9) 1199 (13.5) 1452 (15.9) 1190 (16.6) 791 (14.1) liver disease 60 (1.0) 58 (0.7) 56 (0.6) 54 (0.8) 32 (0.6) renal disease 541 (8.6) 725 (8.1) 797 (8.7) 576 (8.0) 403 (7.2) Functional status 4441 (70.7) 5796 (65.1) 6335 (69.4) 4658 (65.1) 3716 (66.4) requires assistance for transferring 5283 (84.1) 7225 (81.2) 7769 (85.1) 5841 (81.6) 4568 (81.6) requires assistance for dressing 2567 (40.9) 2942 (33.1) 3326 (36.5) 2363 (33.0) 1914 (34.2) requires assistance for eating 4857 (77.3) 6650 (74.7) 7050 (77.3) 5275 (73.7) 4151 (74.2) requires assistance for toileting 5216 (83.0) 7187 (80.8) 7633 (83.7) 5836 (81.5) 4500 (80.4) requires assistance for hygiene 3712 (59.1) 5230 (58.8) 5634 (61.7) 3946 (55.1) 3176 (56.7) bowel incontinence 4914 (78.2) 6799 (76.4) 7195 (78.9) 5464 (76.3) 4240 (75.7) bladder incontinence 4441 (70.7) 5796 (65.1) 6335 (69.4) 4658 (65.1) 3716 (66.4) Devices urinary catheter 153 (2.4) 176 (2.0) 312 (3.4) 194 (2.7) 193 (3.4) dialysis 49 (0.8) 52 (0.6) 39 (0.5) 33 (0.4) 42 (0.8) intravenous medications 14 (0.2) 42 (0.5) 31 (0.4) 25 (0.3) 31 (0.6) tracheostomy 13 (0.2) 26 (0.3) 26 (0.3) 13 (0.2) 10 (0.2) respiratory ventilator 5 (0.1) 10 (0.1) 5 (0.1) 5 (0.1) 5 (0.0) feeding tube 84 (1.3) 89 (1.0) 78 (0.9) 62 (0.8) 55 (1.0) a Data represent number (and percentages) of residents within each quintile. b Median (IQR). diagnostic code for an infectious disease that could help identify the reason for the prescription. However, the three most prevalent antibiotics were agents most commonly used in the context of urinary tract infections (UTIs), including nitrofurantoin (365, 15.4%), trimethoprim/sulfamethoxazole (338, 14.3%) and ciprofloxacin (304, 12.8%). 2860

6 Antibiotic use in nursing homes JAC % of antibiotic treatment courses within facility quintile Lincosamides Metronidazole Second/third-generation cephalosporins Macrolides Third-generation fluoroquinolones First-generation cephalosporins Penicillins Tetracyclines Second-generation fluoroquinolones Sulphonamides Nitrofurantoin Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q Duration of antibiotic treatment (days) stratified by facility quintile Figure 2. Duration of antibiotic therapy across long-term care facility quintiles. The duration of antibiotic treatments among Ontario long-term care residents within facilities stratified into five quintiles from lowest antibiotic (Q1) to highest antibiotic prevalence (Q5). The coloured bars indicate different antibiotic classes and subclasses. Chronic therapy ( 90 days) usually involves nitrofurantoin, sulphonamides or second-generation fluoroquinolones and is most common in Q5. Duration of antibiotic therapy among long-term care recipients The median duration of therapy was 10 days, but there was a wide distribution of prescribed treatment durations (interquartile range 7 30 days) (Figure 1). The most common durations of treatment were 7 days in 691 (29.2%) treatment courses, 10 days in 434 (18.3%) treatment courses, and 14 days in 284 (12.0%) treatment courses. There were also large numbers of residents (496, 21.0%) with antibiotic treatment durations that exceeded 90 days, suggesting chronic prophylaxis or treatment courses. Nitrofurantoin, trimethoprim/sulfamethoxazole and second-generation fluoroquinolones (ciprofloxacin, norfloxacin) were the most common agents used in chronic courses. Variability of antibiotic use across long-term care facilities Long-term care facilities with at least 20 residents (326 of the 363 facilities) were assigned to quintiles based on the facility s point prevalence of antibiotic use. Residents appeared similar across these quintiles with respect to demographic characteristics, healthcare utilization statistics, co-morbidity prevalence, functional status and device use (Table 3). Yet, there was a nearly 5-fold difference in antibiotic prevalence among residents in the lowest quintile ( % receiving antibiotics) as compared with the highest quintile ( % receiving antibiotics) (Table 3). Adjusted analyses demonstrated that in comparison with residents in the lowest prescribing quintile, antibiotic use was more likely for residents in quintile 2 [adjusted odds ratio (AOR) 2.0, 95% confidence interval (CI) ], quintile 3 (AOR 2.7, 95% CI ), quintile 4 (AOR 3.7, 95% CI ) and quintile 5 (AOR 5.5, 95% CI ). Additional analyses were performed to determine whether variations in treatment duration were driving variations in antibiotic point prevalence. Durations of prescribed antibiotic treatment were longer in higher-use quintiles (Figure 2), and this was particularly evident for chronic therapy with antibiotics usually used for urinary infections. Long-course therapy ( 10 days) was provided to the majority of antibiotic recipients (1482, 62.6%), and this proportion increased from the lowest- to highest-use quintile: 60.0% in quintile 1, 58.0% in quintile 2, 59.2% in quintile 3, 63.5% in quintile 4 and 67.7% in quintile 5 (P¼0.0006). If all patients had received short-course therapy (,10 days), only 1606 residents would have been receiving prescriptions on their assessment date, and the point-prevalence estimate would have been only 4.3%. Discussion This population-based, point-prevalence study suggests that antibiotic use is common among long-term care residents, variable across institutions, and may benefit from focused antimicrobial stewardship interventions to standardize treatment indications and duration. 2861

7 Daneman et al. Antibiotic prescriptions were prevalent, with nearly 6% of long-term care residents receiving a course of treatment on the day of their assessment. This prevalence is in line with the 8% point prevalence detected in an evaluation of 4165 residents of 53 Maryland nursing homes, 16 as well as one larger, random sample of residents from 1174 facilities across the USA (point prevalence 10%). 17 The current study provided richer data regarding the types and durations of antibiotic therapy and their contribution to variability across facilities. Nitrofurantoin was the single most commonly prescribed antibiotic among Ontario long-term care residents, accounting for more than 15% of antibiotic prescriptions in this population. This is surprising given that nitrofurantoin is the only antibiotic agent listed among the Beers criteria as a medication that should generally be avoided in elderly individuals. 18 In the updated Beers criteria, nitrofurantoin is listed as a high severity risk, based on the potential for renal impairment and given that there are safer alternatives available. 18 Nitrofurantoin is not actually nephrotoxic. However, it is excreted by the kidneys, and among patients with reduced glomerular filtration rates there may be insufficient active drug in the urinary tract, as well as increased risk of non-renal toxicities (e.g. peripheral neuropathy). It is important to consider the possibility of renal impairment prior to prescribing nitrofurantoin to elderly patients, but strong arguments have been raised that nitrofurantoin ought to be removed from the Beers list. 19 In fact, nitrofurantoin has recently been promoted as a first-line agent for UTIs by the IDSA, due to reduced concerns of promotion of collateral antibiotic resistance as compared with alternative agents. 20 It is difficult to definitively list the indications for antibiotic treatment in our cohort because less than half of these residents were visited by physicians around the time of their prescription, and an even smaller minority had a documented physician claim for an infection diagnosis. This is in keeping with two small case series in the nursing home literature that suggest that most antibiotic prescriptions are called in by telephone and initiated without a recorded physical examination. 16,21 However, the top three antibiotics among Ontario long-term care residents (nitrofurantoin, trimethoprim/sulfamethoxazole and ciprofloxacin) are all used primarily for UTIs, and so this would seem to be the leading indication in our cohort. There is growing evidence that the majority of positive urine cultures are obtained from asymptomatic individuals, and there is no benefit of antibiotic treatment for individuals with bacteriuria in the absence of symptoms (urinary colonization). 22,23 For example, a recent chart review in two Florida nursing homes revealed that 73% of urinary antibiotic treatment courses were inappropriate because the residents did not meet clinical criteria for symptomatic UTI. 24,25 Regardless of whether residents meet appropriate indications to initiate antibiotic treatment, it appears that most are continued on more prolonged therapy than required. Randomized controlled trials (and subsequent meta-analyses) have demonstrated that short-course antibiotic therapy of,1 week is sufficient for common bacterial infections, including cystitis, 26 pneumonia 27 and cellulitis. 28 Yet, nearly two-thirds of antibiotic recipients in our long-term care cohort received treatment courses of 10 days. If all patients had received short-course therapy (,10 days), the estimated point prevalence would have been reduced by 25% (to 4.3%). Even more concerning 10,16,29 31 Finally, are high rates of chronic urinary anti-infective therapy, which suggests prophylaxis against infection. This practice does not have an evidence basis among institutionalized patients, and will inevitably select for resistant organisms given the incredible diversity of the gut microflora. 23 Antibiotic utilization appears to vary markedly across longterm care facilities, with a 5-fold higher prevalence of prescriptions in the highest-use quintile as compared with the lowest-use quintile. This variation does not appear to be driven by resident demographics, history of healthcare utilization, comorbidities, functional status and device dependence. Although random variation is one likely contributor, at least part of this variation may relate to differences in prescriber practices. This notion is supported by longer treatment durations in the higheruse facilities, which should be more dependent on prescriber than recipient characteristics. This is not surprising given that clinical care in each of Ontario s long-term care facilities is usually provided by a single physician. Although this pattern of care may contribute to divergent prescribing practices, it may also facilitate antibiotic stewardship interventions (such as pharmacist review and feedback). Our study has important limitations inherent to the use of retrospective administrative databases, but the use of an independent, comprehensive and accurate drug database has minimized the risk of misclassifying antibiotic use. We elected to measure antibiotic point prevalence to maximize study validity (by ensuring accurate assignment of place of residence at the time of prescription), but this may under-emphasize the extent of antibiotic overuse as compared with longitudinal cohort studies, which typically demonstrate that two-thirds of residents receive antibiotics over a 1 year follow-up period. we cannot rule out that some of these differences may reflect random variation. All of these limitations, though, are outweighed by study strengths, including sample size (largest study to date), use of population-based data (not prone to selection bias of voluntarily participating institutions) and the use of well-validated, rich datasets. Antibiotic use is prevalent among long-term care facility residents and variable across institutions. Long-term care facilities would benefit from antibiotic stewardship programmes to optimize antimicrobial treatment practices, and at the very least should conduct regular reviews of antibiotic utilization. Potentially high-yield stewardship intervention would involve standardizing indications for treatment of UTIs, limiting chronic antibiotic prophylaxis for UTIs and promoting short-course therapy for common bacterial infections. Funding This work was supported by an Interdisciplinary Capacity Enhancement Grant (HOA-80075) from the Canadian Institutes of Health Research (CIHR) Institute of Gender and Health and the CIHR Institute of Aging. N. D. is supported by a CIHR clinician scientist award. C. M. B. is supported by a CIHR and Canadian Patient Safety Institute chair in Patient Safety and Continuity of Care. N. D. and C. M. B. had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. This study was conducted at the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The opinions, results and conclusions reported in this 2862

8 Antibiotic use in nursing homes JAC paper are those of the authors and are independent from the funding sources. S. E. B. is supported by a CIHR new investigator award in the area of ageing. A. G. is supported by a Career Scientist Award from the Ontario Ministry of Health and Long-Term Care. Transparency declarations None to declare. Disclaimer No endorsement by ICES or the Ontario MOHLTC is intended or should be inferred. References 1 Dellit TH, Owens RC, McGowan JE Jr et al. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44: Hecker MT, Aron DC, Patel NP et al. Unnecessary use of antimicrobials in hospitalized patients: current patterns of misuse with an emphasis on the antianaerobic spectrum of activity. Arch Intern Med 2003; 163: Hughes JM. Preserving the lifesaving power of antimicrobial agents. JAMA 2011; 305: Pakyz AL, MacDougall C, Oinonen M et al. Trends in antibacterial use in US academic health centers: 2002 to Arch Intern Med 2008; 168: Ansari F, Erntell M, Goossens H et al. The European surveillance of antimicrobial consumption (ESAC) point-prevalence survey of antibacterial use in 20 European hospitals in Clin Infect Dis 2009; 49: Neuhauser MM, Weinstein RA, Rydman R et al. Antibiotic resistance among gram-negative bacilli in US intensive care units: implications for fluoroquinolone use. JAMA 2003; 289: Edwards JR, Peterson KD, Andrus ML et al. National Healthcare Safety Network (NHSN) Report, data summary for 2006 through 2007, issued November Am J Infect Control 2008; 36: Wenzel RP. The antibiotic pipeline challenges, costs, and values. N Engl J Med 2004; 351: Spellberg B, Guidos R, Gilbert D et al. The epidemic of antibiotic-resistant infections: a call to action for the medical community from the Infectious Diseases Society of America. Clin Infect Dis 2008; 46: Nicolle LE, Bentley DW, Garibaldi R et al. Antimicrobial use in long-term-care facilities. SHEA Long-Term-Care Committee. Infect Control Hosp Epidemiol 2000; 21: Gruneir A, Bell CM, Bronskill SE et al. Frequency and pattern of emergency department visits by long-term care residents a population-based study. J Am Geriatr Soc 2010; 58: Rochon PA, Tu JV, Anderson GM et al. Rate of heart failure and 1-year survival for older people receiving low-dose beta-blocker therapy after myocardial infarction. Lancet 2000; 356: Rochon PA, Stukel TA, Bronskill SE et al. Variation in nursing home antipsychotic prescribing rates. Arch Intern Med 2007; 167: Mor V. A comprehensive clinical assessment tool to inform policy and practice: applications of the minimum data set. Med Care 2004; 42: III Levy AR, O Brien BJ, Sellors C et al. Coding accuracy of administrative drug claims in the Ontario Drug Benefit database. Can J Clin Pharmacol 2003; 10: Warren JW, Palumbo FB, Fitterman L et al. Incidence and characteristics of antibiotic use in aged nursing home patients. JAm Geriatr Soc 1991; 39: Pakyz AL, Dwyer LL. Prevalence of antimicrobial use among United States nursing home residents: results from a national survey. Infect Control Hosp Epidemiol 2010; 31: Fick DM, Cooper JW, Wade WE et al. Updating the Beers criteria for potentially inappropriate medication use in older adults: results of a US consensus panel of experts. Arch Intern Med 2003; 163: Kunin CM. Inappropriate medication use in older adults: does nitrofurantoin belong on the list for the reasons stated? Arch Intern Med 2004; 164: Gupta K, Hooton TM, Naber KG et al. International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: a 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases. Clin Infect Dis 2011; 52: e Katz PR, Beam TR Jr, Brand F et al. Antibiotic use in the nursing home. Physician practice patterns. Arch Intern Med 1990; 150: Nicolle LE, Bradley S, Colgan R et al. Infectious Diseases Society of America guidelines for the diagnosis and treatment of asymptomatic bacteriuria in adults. Clin Infect Dis 2005; 40: Eckburg PB, Bik EM, Bernstein CN et al. Diversity of the human intestinal microbial flora. Science 2005; 308: McGeer A, Campbell B, Emori TG et al. Definitions of infection for surveillance in long-term care facilities. Am J Infect Control 1991; 19: Rotjanapan P, Dosa D, Thomas KS. Potentially inappropriate treatment of urinary tract infections in two Rhode Island nursing homes. Arch Intern Med 2011; 171: Lutters M, Vogt-Ferrier NB. Antibiotic duration for treating uncomplicated, symptomatic lower urinary tract infections in elderly women. Cochrane Database Syst Rev 2008; issue 3: CD Dimopoulos G, Matthaiou DK, Karageorgopoulos DE et al. Shortversus long-course antibacterial therapy for community-acquired pneumonia: a meta-analysis. Drugs 2008; 68: Hepburn MJ, Dooley DP, Skidmore PJ et al. Comparison of short-course (5 days) and standard (10 days) treatment for uncomplicated cellulitis. Arch Intern Med 2004; 164: D Agata E, Mitchell SL. Patterns of antimicrobial use among nursing home residents with advanced dementia. Arch Intern Med 2008; 168: Loeb M, Simor AE, Landry L et al. Antibiotic use in Ontario facilities that provide chronic care. J Gen Intern Med 2001; 16: Benoit SR, Nsa W, Richards CL et al. Factors associated with antimicrobial use in nursing homes: a multilevel model. J Am Geriatr Soc 2008; 56: