|Year : 2015 | Volume
| Issue : 2 | Page : 255-259
Impact of informational feedback to clinicians on antibiotic-prescribing rates in a tertiary care hospital in Delhi
C Wattal1, N Goel1, S Khanna1, SP Byotra2, R Laxminarayan3, A Easton4
1 Department of Clinical Microbiology and Immunology, Sir Ganga Ram Hospital, New Delhi, India
2 Department of Medicine, Sir Ganga Ram Hospital, New Delhi, India
3 Department of Research and Policy, Director, Center for Disease Dynamics, Economics and Policy, Washington, USA; Vice President, Public Health Foundation of India, India
4 Department of Infectious Disease Epidemiology, PhD Candidate, Imperial College London, London, United Kingdom
|Date of Submission||24-Apr-2014|
|Date of Acceptance||09-Sep-2014|
|Date of Web Publication||10-Apr-2015|
Department of Clinical Microbiology and Immunology, Sir Ganga Ram Hospital, New Delhi
Source of Support: The study was financially supported by Center
for Disease Dynamics, Economics & Policy, Washington, DC, Conflict of Interest: None
Context: Antimicrobial use has been associated with increasing antimicrobial resistance. There is an urgent need for judicious use of antimicrobials. Informational feedback has been shown to result in changes in behavioural practices of physicians in certain healthcare settings. We conducted this study to see if the passive informational feedback can reduce in antimicrobial usage in a tertiary care centre. Aims: The study was undertaken to evaluate if the feedback to clinicians on their own antibiotic prescription results in any change in their antibiotic prescription habits. Settings and Design: The study was conducted at a tertiary care setting involving 33 units of different specialties. These units were split into 10 groups based on specialty and were allocated randomly to the control (16 units) and intervention (17 units) arms of the study. This study was a prospective intervention to assess the effect of prescribing feedback on clinical prescribing practices. Materials and Methods: In the intervention arm, information on resistance rates and antibiotic-prescribing patterns was provided to all doctors. Behavioural change was assessed by comparing baseline prescribing rates of each unit with prescribing rates after the intervention. In the control arm, only information on monthly resistance rates was provided. Statistical Analysis : Change in the antimicrobial prescribing rates in the treatment group was assessed by using a Student's t-test. Results: The mean antibiotic use for all the specialties was 189 DDDs/100BDs. The prospective intervention did not elicit any effect on the antibiotic prescribing practices of the physicians. Low prescribers continued to prescribe antibiotics at a low rate, and high prescribers continued to prescribe at a high rate. Conclusions: In view of unfavourable results of passive intervention in the above study, active intervention may be more effective.
Keywords: Antimicrobial-prescription, feedback, intervention
|How to cite this article:|
Wattal C, Goel N, Khanna S, Byotra S P, Laxminarayan R, Easton A. Impact of informational feedback to clinicians on antibiotic-prescribing rates in a tertiary care hospital in Delhi. Indian J Med Microbiol 2015;33:255-9
|How to cite this URL:|
Wattal C, Goel N, Khanna S, Byotra S P, Laxminarayan R, Easton A. Impact of informational feedback to clinicians on antibiotic-prescribing rates in a tertiary care hospital in Delhi. Indian J Med Microbiol [serial online] 2015 [cited 2020 Aug 5];33:255-9. Available from: http://www.ijmm.org/text.asp?2015/33/2/255/153582
| ~ Introduction|| |
Hospitals worldwide face an upsurge in anti-microbial-resistant bacteria causing nosocomial infections.  This has been both a cause and consequence of increased antimicrobial use in hospital settings. Antibiotic use has shown to be an important driver of resistance. , Studies have shown that greater awareness among clinicians of the need for rational use of antibiotics can slow the emergence and spread of multidrug-resistant-organisms (MDRO). 
Various strategies have been attempted in the past to rationalize antibiotic use. For example, patient education has helped reduce antibiotic use by 40% for adults with acute bronchitis, more than what could be achieved by only physician-directed efforts.  In Finland, nationwide recommendations that called for reductions in the use of macrolide antibiotics for respiratory and skin infections in outpatients were issued in 1991 in response to increasing macrolide resistance in group A Streptococcus. A significant decline was found in the frequency of erythromycin resistance among group A streptococci isolated from throat swabs and pus samples.  A study of cardiac surgeons operating in Pennsylvania evaluated the impact on mortality of report cards that provided them with information on their performance relative to their peers. This information led to a decrease in mortality due to increased awareness and intrinsic motivation to provide quality care.  It remains to be seen if similar feedback interventions could be useful in other settings, such as antibiotic prescribing in a hospital.
This study was undertaken as a pilot project to evaluate if the feedback to clinicians of their own antibiotic consumption data results in any change in their antibiotic prescription habits.
| ~ Materials and Methods|| |
We designed a prospective intervention study to assess the effect of prescribing feedback on clinical prescribing practices.
The study was conducted at a 675-bedded tertiary care centre in New Delhi.
Clinicians at our hospital were grouped into units based on their specialties. There are 86 units in our hospital, each made up of one to seven doctors. Among the 86 units, 33 units prescribe about 80% of the total antibiotics prescribed in the hospital and these units were included in the study. Excluded units do not prescribe antibiotics because of their specialty (psychiatrists, doctors practicing alternative medicine, etc.), see very few patients, or prescribe antibiotics infrequently because they are not recommended in most circumstances. The 33 included units were divided into 10 groups based on specialty: Cardiology, chest medicine, general surgery, neurosurgery, obstetrics and gynaecology, medicine, urology, neurology, orthopaedics, and plastic surgery. Within each of the 10 specialty groupings included in the analysis, units were allocated randomly, in approximately equal numbers, to the control (16 units) and experimental (17 units) arms of the study.
The intervention was conducted at our hospital between August 2010 and June 2011. Pre-intervention data from 33 units from July 2009 to June 2010 was obtained and compared with the intervention period for both groups.
In the intervention arm, information on resistance rates and antibiotic-prescribing patterns were provided to all doctors. Antibiotic consumption (in grams) and resistance of organisms were obtained through customized software, Speedminer, (Petaling Jaya, Malaysia), which extracted the data from the Hospital Information System (HIS) (Cambridge, MA, USA). Further, antimicrobial consumption in Daily Defined Dose/100 Bed Days (DDD/100 BDs) was calculated as per the anatomic therapeutic chemical classification (ATC) classification using ABC calc software (available free online). 
Comparison information with peer units within the specialty (without disclosing the identity of the units) and the rest of the hospital was also distributed. In the control arm, only information on resistance rates was provided every month. [Figure 1] shows an example of the feedback that was received by doctors in the intervention group.
|Figure 1: An example of the feedback sheets received by doctors showing the distribution of antibiotic-prescribing rates in the hospital. The dark grey rectangle represents recipient's own unit, the black rectangles represent other units in recipient's specialty, and the light grey area represents all units in other specialties|
Click here to view
Significant behaviour change (P < 0.5) in the treatment group was assessed by comparing baseline (pre-intervention) antibiotic-prescribing rates with post-intervention antibiotic-prescribing rates. These rates were determined using a Student's t-test.
Ethical approval (reference no . EC/03/10/110 ) was obtained from institutional review board before the onset of the study. Data from each unit was anonymous and kept confidential.
| ~ Results|| |
Overall, the average rate of antibiotic consumption between units varied from 110 to 239 DDDs/100 bed-days and 132 to 282 DDDs/100 bed-days in the treatment group and control group, respectively. The mean antibiotic use for all the specialties was 189 DDDs/100BDs [Table 1]. Antibiotic use did not alter meaningfully (P = 0.59) in the treatment arm following the intervention [Table 2]. Surprisingly, the control group showed a significant fall in antimicrobial consumption (P = 0.016). When we reanalyzed the data for this unexpected finding, it was revealed that one of the units in control group that was high prescriber of antibiotics (average antibiotic prescription rate of 379 DDDs/100 BDs) changed to low prescriber of antibiotics (average 123 DDDs/100 BDs). This dramatic fall was due to change in the head of this particular unit coinciding with the beginning of this project. This resulted in change in the antibiotic policy for this unit. However, no other such change of physicians was noted in both control and intervention group. Monthly average DDD/100 bed-days were standardized against the same figures for the previous year in order to account for potential seasonality [Figure 2]. However, the standardized prescribing rates also appeared to be without much variation throughout the intervention. In order to test whether a harmful "boomerang effect" was being observed-whereby doctors who were told they were prescribing below the average increased their prescribing to fall in line with the norm-we separated all units into high and low prescribers [Table 2]. These were defined as units that prescribed above or below the average rate (189 DDDs/100 BDs) in the full year prior to the intervention. Low prescribers continued to prescribe antibiotics at a low rate, and high prescribers continued to prescribe at a high rate [Figure 3]. In our study, we observed that overall (both in treatment and control units) units in cardiology (124-157 DDDs/100 BDs), urology (164-175 DDDs/100 BDs) and plastic and cosmetic surgery (157-169 DDDs/100 BDs) tended to be low prescribers of antibiotics, whereas units in surgery (189-249 DDDs/100 BDs), chest medicine (202-289 DDDs/100BDs) and medicine (181-230 DDDs/100 BDs) tended to be high prescribers of antibiotics [Figure 4].
|Figure 2: Changes in the prescription habits of control and intervention group. DDDs/100 BDs in a given month of 2011 divided by the DDDs per 100 BDs in the same month of the previous year|
Click here to view
|Figure 3: Average DDDs/100 bed-days for units in the treatment group prescribing antibiotics above and below the median during the pre-intervention year.|
Click here to view
|Figure 4: Average antibiotic prescription (DDDs/100BDs) of various units (n = 33) . 1. Cardiology A, 2. Cardiology B, 3. Cardiology C, 4. Chest Medicine A, 5. Chest Medicine B, 6. General Surgery A , 7. General Surgery B, 8. General Surgery C, 9. Superspeciality Surgery D, 10. Medicine A, 11. Medicine B, 12. Medicine C, 13. Medicine D, 14. Neurology A, 15. Neurology B, 16. Neurosurgery A, 17. Neurosurgery B, 18. Neurosurgery C, 19. Neurosurgery D, 20. Obstetrics & Gynaecology A, 21. Obstetrics & Gynaecology B, 22. Obstetrics & Gynaecology C, 23. Obstetrics & Gynaecology D, 24. Orthopaedics A, 25. Orthopaedics B, 26. Orthopaedics C, 27. Orthopaedics D, 28. Orthopaedics E, 29. Orthopaedic Spine Unit F, 30. Plastic & Cosmetic Surgery A, 31. Plastic & Cosmetic Surgery B, 32. Urology A, 33. Urology B|
Click here to view
|Table 1: Antibiotic prescription data of various units (n=33) before and after intervention |
Click here to view
|Table 2: Antibiotic prescription data of treatment (n=17) and control groups (n=16) before and after intervention |
Click here to view
| ~ Discussion|| |
The growing problem of antibiotic resistance in hospitals has prompted many measures to enhance antimicrobial stewardship. Fraser et al. in a randomized trial of an intervention through feedback has shown significant reduction in inappropriate antibiotic prescription without altering the clinical outcome.  Similarly, in another study by Solomon et al. there was reduction in duration of inappropriate antibiotics by 40% in the informational intervention group.  Such studies seldom examine resource poor settings, and this study is the first from India on the effectiveness of intervention program through feedback to the physicians of their own prescription habits in a hospital setting.
In our study, providing doctors with information on how their antibiotic-prescribing rates compared to those of their peers in the present format showed no change in prescribing patterns, even amongst high prescribers. High prescribers, in our study, continued high usage of antibiotics post intervention, possibly due to the nature of the diseases that they were accustomed to treating or lack of meaningful continuing medical education programme targeting antibiotic use. Ours is a tertiary care health centre and therefore has a larger proportion of critical patients, who need more antibiotics due to a higher case mix index (CMI) in certain specialties. CMI is an economic parameter that is calculated using diagnosis related groups, a measure used in various countries as a basis for hospital reimbursement.  Kuster et al. found a significant correlation between antibiotic use and CMI of different departments and units at a tertiary care university hospital.  They showed that antibiotic use varied from 21 in the rheumatology unit, with a lower CMI, to 323 DDD/100 patient-days in the transplantation unit, with a higher CMI. Moreover, there are no references for baseline antibiotic consumption, which can be termed as ideal. High or low antibiotic consumption may be subjective. Antibiotic consumption varies across hospitals based on geographic location, prevalence of MDRO, case mix of patients in particular institutes and availability and affordability of antibiotics. The other reasons for the lack of behavioural change in the antibiotic-prescribing practice of the doctors could be that the doctors did not understand or agree with the message of the information provided or due to the passive nature of intervention used in this study. It has been shown that prospective audit of antimicrobial use with direct interaction and feedback to the prescriber can result in reduced inappropriate use of antimicrobials. 
Studies with negative findings are seldom reported, as there is publication bias towards studies with positive findings.  However, such studies can also significantly contribute to the lacunae in knowledge. Currently, there is scarce data on antimicrobial prescription habits of physicians from India. The strength of our study is that it has highlighted for the first time the antimicrobial consumption patterns of broad range of specialties over a period of 23 months from India. The result of our study also stresses that since passive intervention did not elicit desirable behavioural change in the physicians, the possibility of direct interaction with the prescribers to reduce antimicrobial consumption may be more effective, at least in our setting. These methods may include training workshops, focus group discussions, coordinated implementation of antibiotic policy, preauthorization of specific antibiotics, de-escalation of therapy and computer-assisted prescription strategies. ,
| ~ References|| |
Friedland I, Stinson L, Ikaiddi M, Harm S, Woods GL. Phenotypic antimicrobial resistance patterns in Pseudomonas aeruginosa
and Acinetobacter: Results of a multicenter intensive care unit surveillance study, 1995-2000. Diagn Microbiol Infect Dis 2003;45:245-50.
Seppälä H, Klaukka T, Vuopio-Varkila J, Muotiala A, Helenius H, Lager K, et al
. The effect of changes in the consumption of macrolide antibiotics on erythromycin resistance in group A streptococci in Finland. Finnish Study Group for Antimicrobial Resistance. N Engl J Med 1997;337:441-6
Iosifidis E, Antachopoulos C, Tsivitanidou M, Katragkou A, Farmaki E, Tsiakou M, et al
. Differential correlation between rates of antimicrobial drug consumption and prevalence of antimicrobial resistance in a tertiary care hospital in Greece. Infect Control Hosp Epidemiol 2008;29:615-22.
Nava JM, Bella F, Garau J, Lite J, Morera MA, Martí C, et al
. Predictive factors for invasive disease due to penicillin-resistant Streptococcus pneumoniae
: A population-based study. Clin Infect Dis 1994;19:884-90.
Gonzales R, Corbett KK, Leeman-Castillo BA, Glazner J, Erbacher K, Darr CA, et al
. The "minimizing antibiotic resistance in colorado" project: Impact of patient education in improving antibiotic use in private office practices" Health Serv Res 2005;40:101-16.
Fraser GL, Stogsdill P, Dickens JD Jr, Wennberg DE, Smith RP Jr, Prato BS. Antibiotic optimization. An evaluation of patient safety and economic outcomes. Arch Intern Med 1997;157:1689-94.
Solomon DH, Van Houten L, Glynn RJ, Baden L, Curtis K, Schrager H, et al
. Academic detailing to improve use of broad-spectrum antibiotics at an academic medical center. Arch Intern Med 2001;161:1897-902.
Kuster SP, Ruef C, Bollinger AK, Ledergerber B, Hintermann A, Deplazes C, et al
. Correlation between case mix index and antibiotic use in hospitals. J Antimicrob Chemother 2008;62:837-42.
Dellit TH, Owens RC, McGowan JE Jr, Gerding DN, Weinstein RA, Burke JP, 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;44:159-77.
MacDougall C, Polk RE. Antimicrobial stewardship programs in health care systems. Clin Microbiol Rev 2005;18:638-56.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]