Evaluation and optimization of antibiotic usage in upper respiratory tract infections in children at a tertiary care outpatient department: A clinical audit.

INTRODUCTION: Inappropriate antibiotic (ab)use contributes to antimicrobial resistance. Upper respiratory tract infection (URTI) is the most common reason for antibiotic prescription in an outpatient department (OPD). Several factors influence the high and unjustified antibiotic use in a common ailment.
MATERIALS AND METHODS: A clinical audit was performed to assess antibiotic prescription rate (APR) for URTI in the pediatric OPD against the available benchmark. The prescription pattern was assessed, and interventions were formulated to improve prescription behavior. Data of all children attending OPD and fulfilling the criteria for URTI group were collected from the online hospital management system and analyzed. Interventions, in the form of discussions, presentations, posters, and guidelines (Indian Ministry of Health Guidelines for URTI) regarding etiology of URTI, and indications for antibiotic prescription were implemented. Data were monitored and feedback to consultants was given.
RESULTS: The baseline APR was 14.7%. There was wide variation in APR (4.1%-53.1%) among consultants. Three consultants had a rate of 53.1%, 29.7%, and 28.6%, which was very high. Postintervention, the average APR decreased to 8.7%, a reduction of 40.8%. There was a reduction in APR among consultants with high APR as well. There was reduction in the use of azithromycin, a drug recommended for patients with penicillin allergy, from 21.2% to 14.4% (32.1% reduction). Amoxycillin plus clavulanic acid combination and amoxicillin alone continued to be the most prescribed antibiotics.
CONCLUSION: Interventions through clinical audit were useful in reducing APR. The APR of 8.7% achieved in this study postintervention can be used as a benchmark by other institutions to assess APR in children with URTI.

Introduction

Self-medication is rampant.[12] Indian society is no exception. One of the risks of self-medication is unsupervised and inappropriate antibiotic (ab)use, which may contribute to antimicrobial resistance.[3] Continuing medication on an old prescription for an acute illness without medical practitioner's advice has emerged as a common form of self-medication and is used for adults as well as for children. Since the original prescription was meant for a single episode of illness, repeat administration of the same medication for similar symptoms may seem justified. However, similar symptoms do not mean the same disease. In such situation, it may be dangerous to self-medicate on the same prescription as it may lead to delay in diagnosis and appropriate treatment. At the same time, it may be responsible for the use of antibiotics where none was necessary. Antibiotic use is associated with increased risk of isolation of antibiotic-resistant organisms.[4] Prescription behavior of general practitioners and over-the-counter dispensing of medicines by pharmacists often mimic that of the specialists.[5] Therefore, it becomes necessary that prime institutions lead the way in prescription behavior.

There are many studies that show high antibiotic prescription rate (APR) in upper respiratory tract infection (URTI) and wide variation in antibiotic use by clinicians.[567] There is also wide variation in the choice of antimicrobials prescribed.[7] Various reasons are attributed to such behavior, and they range from commercial interests to lack of knowledge, time, and parental pressure.[78]

Materials and Methods

The primary objective of the current study was to assess the APR in pediatric outpatient department (OPD) for URTI patients with respect to available benchmark and to decide on interventions required to improve the prescription behavior.

The secondary objective was to assess the antibiotic prescription pattern for URTI in our hospital.

The study, a clinical audit, was conducted in the pediatric OPD of our hospital, between June 1, 2019, and March 31, 2020. Our hospital is a 983-bedded tertiary care hospital, owned by a steel giant, and provides free treatment to its employees and their family members and, on a payment basis, to the community at large. Paying patients form approximately 25% of the OPD attendance in the pediatric OPD.

The hospital management system (HMS) has an OPD module which is used to generate on-line prescriptions for patients entitled for free treatment which are dispensed at the in-house pharmacy counters. For paying patients, hard copy of prescription is given, and they buy medicines from outside the hospital. This prescription is not entered in HMS to avoid duplication of work. The diagnosis in online OPD form is not coded as per the ICD and is written as free text by the doctors. This project was registered as a self-initiated project on total quality management website of our organization, with due ethical clearance.

Literature search for benchmark on APR for URTI did not reveal clear-cut guidelines. Individual reports suggested that 60%–75% infections were viral.[9] A Cochrane review suggested that antibiotics were used in delayed antibiotic group in 27%–38% of patients based on the method of prescription filling, i.e., handing over delayed prescription versus keeping prescription ready at the primary care center.[10] Based on this, an acceptable level of antibiotic use was defined as <25% of total URTI treated. Guidelines on antibiotic use suggested that amoxicillin, co-amoxiclav, and azithromycin, in case of penicillin allergy, were appropriate first-line antibiotics.[1112] Thus, these two antibiotics were considered appropriate antibiotics for the treatment of URTI. Azithromycin was considered appropriate if penicillin allergy was documented.

The following data for the full financial year 2018–2019 (FY-19) were generated through HMS: number of patients seen in pediatric OPD, diagnosis of these patients, medicines prescribed, name of the prescribing doctor, and tests ordered. The data were segregated doctor-wise. The doctor's name was coded, and the coded data were given to other authors involved in the study. The person who coded the data (SS) was involved in analysis, but not in the assessment process. Analysis of baseline data was completed in July 2019. August and September 2019 were used to decide and implement the interventions. Postintervention data were collected from October 2019 to March 2020. The same method was repeated while analyzing data in the postintervention period each month.

For analysis, the following conditions that have overlap of symptoms were included in the URTI group: upper respiratory infection, URTI, common cold, cold, rhinitis, acute rhinitis, sinusitis, rhinosinusitis, pharyngitis, acute otitis media, acute respiratory infections, acute respiratory tract infections, and viral fever. The diagnoses entered as wheeze-associated lower respiratory infections, reactive airway disease, asthma, bronchitis, and lower respiratory infections were not included in the study.

Master sheet for baseline data was created in Excel with the following noted against the patient name: presence of URTI (Y for yes), antimicrobial prescription given (Y for yes), and name of antibiotic. Numbers of URTI group patients, those receiving antimicrobials, and type of antimicrobials were thus counted. This was the baseline data.

After analyzing baseline data, interventions for improvement were planned through brainstorming session involving all doctors in the department of pediatrics. Variations in antibiotic use and overall data were shared with all the doctors. However, prescription rate of an individual doctor was not shared with doctors at this stage. Initially, a recapitulation session on URTI was organized for doctors alone, where doctors with the highest prescription rate for antibiotics were invited to speak on etiology of URTI, diagnosis of bacterial URTI with indications for antibiotic therapy, and choice of antibiotics. Ministry of Health and Family Welfare guidelines was also reviewed with respect to respiratory infections.[1112]

Next, posters stating that URTI is viral and does not need antibiotics were displayed in OPD consultation chambers. This served as a public education tool as well as a reminder to consultants.

The third intervention was to monitor data on a monthly basis and provide feedback to consultants.

Further intervention planned was documenting reason for antibiotic use in the OPD prescriptions and software modification in the HMS where the prescribing doctor would need to link the prescription to an indication; these were not found practical due to the cost of modification and the time required in a busy OPD.

Statistical analysis

Statistical analysis was done on Microsoft Excel sheet. APR for URTI was defined as number of patients receiving antibiotics/100 patients of URTI group. SPSS 23 software was used to calculate Pearson's correlation coefficient and P value.

Observations

Baseline data showed that 38,135 patients registered for consultation, of which 20,552 patients had adequate information and were included in study. Of these, 7912 (38.5%) patients belonged to URTI group. During the postintervention period of 6 months from October 2019 to March 2020, 21,458 patients registered. Among these, adequate data were available for 12,143 patients and 5817 (48%) patients belonged to URTI group [Tables 1 and 2]. This difference is highly significant (P < 0.0001)

Table 1

Baseline data for the year 2018-2019 (financial year-2019) showing method of coding, audit and results

Audited by doctors	Prescribed by	Total OPD patients	URTI group	AB used	Percentage AB used	AB used*
						Am	CoA	Ce	Az	Of	Cef
A	1	787	196	17	8.7	4	11	0	0	2	0
B	2	2848	831	75	9	15	29	0	24	7	0
D	3	2981	751	72	9.6	30	28	2	4	8	0
A	4	4394	399	212	53.1	1	151	4	54	2	0
A	5	2130	579	37	6.4	0	23	0	14	0	0
D	6	450	98	5	5.1	2	3	0	0	0	0
B	7	2066	721	48	6.7	17	22	0	4	3	2
C	8	4661	886	253	28.6	15	127	4	102	5	0
B	9	3471	893	118	13.2	6	49	27	21	15	0
C	10	4031	586	174	29.7	26	113	17	17	1	0
A	11	1240	212	31	14.6	13	15	0	2	1	0
C	12	3179	569	33	5.8	18	6	6	0	3	0
D	13	2825	640	49	7.7	0	9	37	2	1	0
D	14	1222	87	18	20.7	1	11	3	1	2	0
B	15	1850	464	19	4.1	3	14	0	1	1	0
Total		38,135	7912	1161	14.7	151	611	100	246	51	2

*Am=Amoxycillin, CoA=Co-amoxiclav, Ce=Cephalexin, Az=Azithromycin, Of=Ofloxacin, Cef=Cefuroxime, AB=Antibiotics, OPD=Pediatric outpatient department, URTI=Upper respiratory tract infections

Table 2

Trend in use of antibiotics over study period

	Patients analyzed	URTI group	Percentage URTI group	AB used	Percentage of patients AB used	AB used*
						Am	CoA	Ce	Az	Of	Cef
FY 2019	20,552	7912	23.1	1161	14.7	151	611	100	246	51	2
October 2019	2249	964	24.6	100	10.4	5	60	11	23	1	0
November 2019	2221	1233	33.2	92	7.5	7	55	6	20	3	1
December 2019	2017	1032	30.8	84	8.1	16	45	3	15	3	2
January 2020	1947	679	22.3	78	11.5	15	46	3	11	2	1
February 2020	2135	941	27.3	74	7.9	9	50	12	1	2	0
March 2020	1574	968	24.3	78	8.1	12	50	10	3	2	1
Total 6 months (FY-2020)	12,143	5817	48	506	8.7	64	306	45	73	13	5

*Am=Amoxycillin, CoA=Co-amoxiclav, Ce=Cephalexin, Az=Azithromycin, Of=Ofloxacin, Cef=Cefuroxime, AB=Antibiotics, URTI=Upper respiratory tract infections, FY=Financial year

Baseline data showed that average antimicrobial prescription rate in FY-19 (2018–2019) was 14.7% in URTI [Table 1]. Based on our target of <25% prescription rate of antibiotics, it was well within the acceptable limits and did not require further intervention. However, doctor-wise analysis showed a variation from 4.1% to 53.1% [Table 1]. This wide variation required to be addressed. Three doctors had an APR rate of 53.1%, 29.7%, and 28.6%. Subsequent monitoring of trends focused on total prescription rate and prescription rates of these doctors. Trend for total prescription is shown in Figure 1. Postintervention, the average APR was 8.7% which showed a reduction in APR by 40.8%. The antibiotic prescription trends for three doctors are shown in Figure 2. Both figures show a postintervention downward trend in the prescription rate of antibiotics.

Figure 1

Overall trend of antibiotic prescription for upper respiratory tract infection patients

Figure 2

Trend of Antibiotic prescription for “high prescribers”

Type of antibiotics prescribed [Table 1] were co-amoxiclav (52.6%), amoxicillin (13%), azithromycin (21.2%), cephalexin (8.6%), and others (4.6%). Other drugs were ofloxacin and cefuroxime used occasionally in children suffering from diarrhea along with respiratory infection. In subsequent months, a gradual reduction in the use of azithromycin (recommended for children with penicillin allergy) was seen [Table 2]. Since azithromycin is not being used specifically for penicillin allergy, appropriate APR for first-line antibiotics is 65.6% considering amoxicillin and co-amoxiclav as the first-line antibiotics for URTI. Postintervention, prescription of appropriate first-line antibiotics (amoxicillin and co-amoxiclav) increased to 73.1% which showed a marginal improvement by 11.4%. Azithromycin use decreased significantly from 21.2% to 14.4%, showing a decline by 32% [Table 2].

Data collection was discontinued from April 2020 to October 2020 as our hospital was designated a COVID hospital and the routine OPD was suspended. Therefore, data from April 2020 to October 2020 were not analyzed. As normal OPD restarted in November 2020, we analyzed data for November 2020 and December 2020 to see if the effect of interventions persisted. APR in November and December showed a value of 36 (11.6%) out of 311 and 26 (10.9%) out of 239 cases of URTI, respectively.

Table 3 depicts month-on-month comparison between the pre- and post-intervention month showing a consistent decline in APR.

Table 3

Comparison of antibiotic prescription in same months in 2 study years

Months	Percentage antibiotics prescribed
	FY
	2018–2019 (FY-2019)	2019–2020 (FY-2020)
October	16.1	10.4
November	16.1	7.5
December	14.2	8.1
January	20.4	11.5
February	16.1	7.9
March	Not analyzed	8.1

FY=Financial year

Our analysis also revealed that there is no statistically significant correlation between the number of patients seen and rate of antibiotic prescription as shown by a Pearson's correlation coefficient of 0.395 and P = 0.116. Figure 3 depicts the relationship between number of patients seen by a consultant and APR.

Figure 3

Relationship of total patients seen and antibiotic usage. Statistical analysis shows a Pearson's correlation coefficient 0.395 and P = 0.116, suggesting absence of correlation between number of patients seen and antibiotics prescribed

Discussion

Antimicrobial resistance is an emerging epidemic that calls for action to reduce the development of antibiotic resistance. Antibiotic (ab)use is the prime reason for developing antibiotic resistance.[1314] URTI is the most common reason for OPD prescriptions of antibiotics.[1516] Therefore, URTI was chosen as a subject for this study. Our baseline data for 1 year (2018–2019 or FY-19) suggested that we were within the acceptable range of overall APR (14.7%). However, doctor-wise analysis suggested over-prescription of antibiotics by three doctors that needed correction. The intervention showed that with proper training, reminders, and regular feedback on the APR, the prescription rates of these doctors came down. COVID-19 pandemic interrupted this study as our hospital was designated a COVID hospital, and routine work was scaled down. As regular OPDs resumed in November 2020, we saw a little uptrend in APR (11.6%), which was still lower than the APR of baseline year. This came down further in December 2020 (10.9%) with reinforcement of the feedback, suggesting that regular feedback leads to improvement in appropriateness of prescription.

Postintervention data suggested that the number of patients within URTI group increased significantly from 38.5% to 48%. This appeared to be the result of awareness about monitoring and consequent improved documentation of respiratory problems. This was an unexpected gain from the study.

Another gain from the interventions was reduced use of azithromycin. It is good to reduce the use of this drug so that the development of resistance is delayed or prevented. Azithromycin is particularly important as it is a drug, useful for the treatment of multidrug-resistant typhoid fever.

During the postintervention period, January 2020, data showed a spike in APR [Table 3]. This led us to compare APR month on month. This comparison showed that APR was highest (20.4%) in January 2019, in the FY-19, and there was a reduction in APR in January 2020 as well. Possible reasons for higher APR may be seasonality associated increased URTI rates, higher incidence of bacterial infections due to peak winters, increased travel as schools reopen after winter vacations, and possible parental anxiety to send children to school, leading to increased demand for “quick relief.”[17] Bacteriological studies are required to confirm or refute this hypothesis. A study in Bengaluru on airborne viable bacterial pollution showed that Corynebacterium and Streptococcus were the most common bacterial pollutants in winters, although the highest concentration of bacterial pathogens, including Streptococcus pneumoniae was seen in premonsoon season.[18]

During the study, it was suggested that the doctors who have larger number of patients may be using antibiotics more frequently due to time pressure in busy OPDs. However, Figure 3 clearly shows that there is no correlation between number of patients seen by a doctor and APR. It helps in emphasizing the fact that busier doctors do not prescribe more antibiotics and should be a lesson for the younger trainees in pediatrics.

Limitations of this study include our inability to include paper prescriptions in the analysis. It may be argued that consultants may have a different prescription pattern for people who purchase medicines, and it may influence overall APR. However, considering that such patients account for 25% of OPD attendance, conclusions drawn from our data seem reliable. Another limitation is our inability to link the prescription to clinical signs observed and microbiological investigations. However, this is the reality of OPDs in resource-limited settings, and our study contributes to better understanding of the prescription patterns and ways to reduce APRs.

Conclusion

The present study achieved an APR of 8.7% for URTI. Clinical audit methodology helped us to reduce APR by 40.8% and use of azithromycin by 32% in URTI in children. Based on the observation and analysis of data from this study, we infer that the APR and appropriateness of choice of antibiotics can be enhanced through clinical audit. Monitoring and feedback at regular interval seem the most effective intervention to achieve the goal of appropriate antibiotic prescription for URTI in children. APR of 8.7% can be used as a benchmark by other institutions in resource-limited settings to compare their data and optimize antibiotic use in pediatric OPD in children suffering from URTI as defined in the present study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.