Introduction Infections are a common cause of paediatric morbidity. Antibiotics are vital in treating them. Erratic prescribing practices are an important cause for the development of antibiotic resistance. Our objective was to estimate the effectiveness of educational interventions to improve empirical antibiotic prescribing practices among paediatric trainees. We aimed to improve the compliance to antibiotic protocols and to sustain it over 6 months.
Methods It is a time interrupted non-randomised trial conducted in a tertiary hospital in India. Initially, 200 admitted children were selected randomly. Their antibiotic prescriptions, adherence of prescriptions to the then existing antibiotics guidelines, course during hospital stay and the final outcome were noted. The existing antibiotic policy and its use were reviewed. It was then considered essential to prepare a fresh antibiotic policy based on national guidelines, local sensitivity patterns and with inputs from microbiologists. This was distributed to the residents through seminars, posters and cellphone friendly documents. Compliance to the policy was also tracked twice a week. The adherence to guideline was recorded in the subsequent 6 months.
Results The adherence of empirical antibiotic prescriptions was 59% before intervention which improved to 72% in the first month, 90% in the second month, 86% and 78% in the third and sixth months, respectively. There was no significant difference in duration of stay and the outcome at discharge in the patients in adherent and non-adherent groups.
Conclusion Educational interventions and frequent monitoring improved antibiotic prescribing practices among residents with no negative impact on patient outcomes. Quality improvements need persistent reinforcement and frequent monitoring to be sustainable.
- Antibiotic management
- Quality improvement methodologies
- Staff Development
Data availability statement
Data are available on reasonable request. Analysed data are available in the manuscript.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
Improving antibiotic prescribing is critical to effectively treat infections, protect patients from unnecessary side effects, and combat antibiotic resistance.
WHAT THIS STUDY ADDS
Designing a locally accepted antibiotic policy and educating the doctors is of paramount importance to improve prescribing practices. Our study shows that the adherence increased from 59% before intervention to 78% after intervention at 6 months.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Adherence to guidelines should be monitored and updated regularly to maintain the improvement.
In a developing country such as India, infectious diseases are the most common cause of morbidity and mortality.1 Antibiotics are crucial in treating them. Misuse and overuse of these drugs can lead to emergence of multidrug resistant micro-organisms.
A study done by Kotwani and Holloway found high antibiotic consumption in the capital metropolitan city of India: 43 390 defined daily doses (DDD) per 1000 patients in the public sector, 125 544 DDD per 1000 patients in private pharmacies, and 81 467 DDD per 1000 patients in private clinics.2 Data from the Indian Council of Medical Research (ICMR) antimicrobial resistance surveillance network show that more than 70% Enterobacteriaceae are resistant to third-generation cephalosporins.3 In 2010, India was the world’s largest consumer of antibiotics for human health at 12.9×109 units (10.7 units per person).4
Behaviours related to use of antibiotics need change to decrease the incidence of antimicrobial resistance. India has achieved remarkable reductions in smoking in buildings and workplaces through regulation and behaviour change communications. Similar campaigns could work to educate the public and physicians about the dangers of uncontrolled antibiotic use, as has been the case in high-income countries, but more research is needed to see how well this could work in hospitals in India.5 Worldwide, several methods and interventions have been used to improve the use of antibiotics by physicians including using a checklist for indication of use of antibiotics, educational interventions (seminars, index cards), declaring official antibiotic policy and dissemination of antibiotic policy using the intranet .3 5 6
There may not be a one-for-all method in ensuring adherence to the policy in different hospitals. Every clinical unit would require its own method of implementation and improvement in compliance to it. In 2017, the faculty and residents in our hospital were involved in quality improvement training using point-of-care quality improvement method.6 This motivated us to undertake this study. We decided to review and improve the use of antibiotics in our paediatric medical units.
Our objective was to improve antibiotics prescribing practices of resident doctors working in our unit. We aimed to improve the compliance to antibiotic protocols by 50% and to sustain it over 6 months by implementing an educational programme, supplemented by frequent monitoring and re-emphasis on the compliance.
This quality improvement initiative was undertaken at Kalawati Saran Children’s Hospital attached to Lady Hardinge Medical College in New Delhi, India (LHMC). It has 380 beds with more than 25 000 inpatients and more than 250 000 outpatient visits per year. This study was conducted in one of the three paediatric medical units. All data were collected by anonymising patient identifiable information.
The first prescriptions at admission are written by the residents who are the first point of contact to the patients. They are generally rotated through the different units of the hospital for training purposes (figure 1). The institutional (LHMC) guidelines for antibiotics use were prepared in 2015 and the residents could also access the guidelines issued by ICMR, New Delhi, and by National Center for Disease Control, New Delhi (NCDC) on the internet. We did not have a formal written antibiotics policy of our unit.
The data on antibiotic compliance were collected from 1 November to 31 December 2017 and this was used as a control group . We included children admitted to our unit (from outpatient department and emergency services) between the ages of 0 and 18 years and who received antibiotics at admission for any infectious disease. Children who already had a culture report or had received antimicrobials previously in a different healthcare centre were excluded.
Two hundred cases which met our inclusion criteria were selected from the unit’s admission register using random number tables and their inpatient files were accessed. The provisional diagnoses, the antibiotics prescribed at admission, the changes in treatment during the course of stay, the reasons for change in antibiotics and the final outcome were noted. The prescriptions were considered adherent to guidelines if they adhered to any one of the three guidelines viz the guidelines from LHMC, ICMR or NCDC. These guidelines were selected as they were standardised, updated and accessible to all.
Adherence to either one or more of the existing guidelines was found to be 59% only. In order to improve the adherence, it was decided first to find the reasons behind this inadequate adherence. A multidisciplinary quality improvement team was formed consisting of faculty (paediatrics, neonatology and microbiology), residents and the nursing officers to coordinate this quality improvement initiative. The team formulated a Specific Measurable Achievable Relevant Time-bound (SMART) aim—to increase the adherence of first prescription of antibiotics to the guidelines from current 59% to greater than 90% in 6 months’ time. The team conducted a root-cause analysis to detect the causative factors for non-adherence to the guidelines. A fish bone analysis (figure 2) also illustrated the various possible reasons for non-adherence. The most important and easily remediable reasons we found were that there existed a hospital policy but it was not easily accessible at the workplace where patients came into first contact for prescribing antibiotics and that the policy did not cover all the common illnesses which present to our hospital.
The team decided to create a written policy on the recommended empirical antibiotics for all the common infectious illnesses. While formulating the policy, we found that no single guideline could give us the appropriate empiric antibiotic regimen for all the common diagnoses. So we consulted the guidelines by NCDC and ICMR in addition to the guidelines of our own hospital for designing it. The experts in the field of pulmonology and severe acute malnutrition were also consulted because a considerable fraction of inpatients tend to suffer from severe acute malnutrition and respiratory infections. The culture and sensitivity patterns of common organisms of the previous 5 years (2013–2017) were reviewed and used in finalising the policy.
The policy was then scrutinised by our team and the other faculty in the unit. The policy was also approved by the head, department of paediatrics in our hospital. This empirical antibiotic policy designed and customised for our hospital is given in online supplemental Annexure 1. The primary intervention was to formulate the empirical antibiotic policy specific to our unit which included all the common diagnoses and to introduce it to all the residents through an educational seminar. We planned to display the posters of the policy at prominent places in the hospital ie the paediatric emergency room, doctors’ duty room and treatment rooms in the unit. We also sent a downloadable portable document format (pdf) copy of the policy through social media to each resident (plan-do-study-act, PDSA cycle 1). Our major strategy for implementation was education, auditing, feedback and frequent reminders (figure 3).
The first PDSA cycle consisted of adopting the policy as the policy of the unit, circulation of the policy through a seminar and also to distribute it in the form of pasting the posters in the areas where patients are seen first (ie, emergency room and also in the treatment room of our unit) and pdf documents for easy access and quick reference. In September 2018, a seminar was conducted to introduce the antibiotic policy to all the residents and the faculty. The seminar provided an overview of the guidelines, discussed the rationale behind the guideline recommendations and identified situations where local practice diverged from the national guidelines due to the hospital microbial culture and sensitivity patterns. It also provided the results of the baseline study that we had conducted a few months previously as well as the recent studies conducted around the country regarding the dramatic rise of antimicrobial resistance.
Any new residents posted to the unit had an induction which provided the antibiotic policy and stressed on the importance of following it so as to maintain the effects of implementation.
Following the introduction of antibiotic policy, 50 random cases were selected at the end of first, second, third and sixth months and their adherence to the antibiotic policy was noted in terms of percentage. A χ2 test was applied to know the statistical significance of the change.
After noting the results at the end of first month, the team reviewed the progress. The adherence to the new policy increased, but the change was not statistically significant. A meeting was called at the end of first month and the residents were requested to give their verbal feedback regarding barriers for the implementation. Also their queries, suggestions, situations where they found difficulties in following the policy were reviewed and their possible solutions were also discussed.
We decided to have reinforcements such as weekly reminders to follow the antibiotic policy and the need to follow it was sent in the official online/WhatsApp group for doctors of the unit (PDSA cycle 2).
We started a biweekly review of the antibiotic prescriptions in morning rounds. The reasons for non-adherence were discussed and the residents were instructed and educated for appropriate choice of antibiotics. The adherence to antibiotic policy was reviewed at the end of second, third and sixth months. We also compared the change in antibiotics after admission, and the final outcome at discharge.
The primary outcome was measured as the proportion of prescriptions adherent to the antibiotic policy. Hospital staff of 10 consultants, 64 residents and 24 nurses were targeted to participate in this programme. We planned an educational intervention and since the residents rotated into the unit change every 3 months, this seemed the best way to study the impact of our intervention on the prescribing behaviours of residents.
The median age of the patients in the whole cohort was 12 months (range 2 days to 17 years). Among them 58.75% were males. The cases were evaluated at baseline and followed up till the conclusion of their treatment—either till the child became well and was discharged or succumbed to the clinical consequences.
In the preintervention period, the adherence to guidelines was 59%. After the intervention (first PDSA cycle), at the end of the first month even though the adherence increased to 72% (p=0.09), it was not significant statistically. A review meeting was held to discuss the possible hindrances in implementation of the antibiotic policy. The residents were provided with the evidence regarding the causes of development of antimicrobial resistance, its effects on human health, the ways to tackle it and the necessity of antibiotic policy. This led to a significant rapid change in the adherence to the policy. Following this, the second PDSA cycle was done using significant reinforcement measures in the form of reviewing the adherence in the morning rounds (at least twice weekly on the day following the admission days, ie, Tuesdays and Fridays) and it led to an increase in adherence to 90% (p=0.04). The improvement persisted at the end of third and sixth months as 86% and 78%, respectively (p<0.05). Overall, the postintervention antibiotic compliance was 81.5% (p<0.00001). The run chart of the adherence to antibiotic policy is shown in figure 4.
We also reviewed whether the empirical antibiotics were changed and the reasons for that. We found that the common reasons for change in the antibiotics were non-improvement of clinical status, worsening clinical condition, availability of culture and sensitivity report, in this order. The children in the non-adherent group were more likely than the children in adherent group to have their antibiotics changed during the course of their treatment (figure 5).
On subgroup analysis, we found that the adherence improved drastically postintervention in bronchiolitis, fever without focus and acute gastroenteritis compared with the preintervention period. The change in adherence in pneumonia and sepsis, however, varied in different months.
The diagnosis of pneumonia has never achieved a 100% adherence and the rates of non-adherence were also high in sepsis (figure 6). Bronchiolitis which had a higher percentage of non-adherence before intervention saw a drastic change with >90% adherence in the subsequent phases. Peritonitis, fever without focus, liver abscess and acute gastroenteritis saw a rapid and drastic improvement in their adherence postintervention.
The difference in duration of stay among the adherent and non-adherent groups remained insignificant. It implies that there was no clinical harm or increase in the duration of stay if antibiotics were used according to the unit policy. There was no change in the final outcome (ie, discharge or death) of the patients in both the adherent and non-adherent group which re-emphasises the preceding point of prescribing the antibiotics according to the policy.
The reasons put forward by the residents as the reason for resistance to change their previous prescribing practices were fear of under treatment, over estimation of severity of clinical condition and the feeling that these policies are not ideal in an overcrowded government tertiary care centre. All these challenges were overcome by frequent open discussions and spending some time together analysing the evidence behind why antibiotics are advised in the policy.
Our study is a single-centre, hospital-based, non-randomised controlled trial with an interrupted time-series analysis. Two hundred patients were enrolled in the study to note the baseline data before the intervention. After the introduction of antibiotic policy, 200 cases were again enrolled, 50 each at the first, second, third and sixth months. The root cause analysis revealed deficiencies in the policy and also the lack of accessibility to it. Quality improvement efforts included a new antibiotic policy, wide dissemination and availability at the point-of-care and frequent reinforcements using regular reviews and monitoring. This led to an improvement in adherence to the policy from 59% to 72% in the first month, which later changed to 90% by the end of second month, and 86% and 78% in the third and sixth months, respectively.
Similar to our intervention of making a policy and implementing it with frequent reinforcements, Stocker et al used a mandatory checklist requiring indication and recording likelihood of infection at start of antibiotic therapy, review of the continuing need for therapy at 48 hours and 5 days, and documenting the reasons for continuation and possible target pathogen. The use of appropriate empiric antibiotic therapy courses for culture-negative infection-like symptoms increased from 18% (10/53) to 74% (42/57; p<0.0001), duration of therapy <3 days increased from 18% (10/53) to 35% (20/57; p=0.05) and correct targeting of pathogen increased from 58% (7/12) to 83% (20/24; p=0.21).7
Ambroggio et al in their study found a significant improvement in appropriate first line antibiotic prescribing for uncomplicated pneumonia by the emergency team from 0% to 82% and in the ward team from 30% to 90% over a span of 3 months after a series of interventions including educational seminar, providing summary of guidelines, index card, and changes in electronic medical record prescription system.8
Hadi et al found that after the creation of the hospital’s official antibiotic guideline, the proportion of patients treated with antibiotics decreased from 88% to 54% (effect size –34% points, 95% CI –25% to –43%). The amount of antibiotics used decreased to almost half from 99.8 to 53 DDD/100 patient‐days.9 Thus, having a formal policy made this difference.
Chandy et al found that the hospital antibiotic policy guidelines disseminated through a booklet and intranet access in all outpatient, departmental offices and ward computers resulted in a significant decrease in antibiotic use .4
We also found that the change of antibiotics during the course of treatment was more likely in the non-adherent group than in the adherent group. We could not find other studies to compare these observations.
We found that the final outcome at discharge in both adherent and non-adherent group was similar. A Cochrane review to estimate the effectiveness and safety of interventions to improve antibiotic prescribing to hospital inpatients and to investigate the effects of two intervention functions (restriction and enablement) found that the duration of antibiotic treatment decreased by 1.95 days. It also showed interventions to be associated with improvement in prescribing practices according to antibiotic policy in routine clinical practice. The risk of death was similar between intervention and control groups (11% in both arms), indicating that antibiotic use can be reduced without any increase in mortality.10
Though the most appropriate techniques for quality improvements in antibiotic use in paediatric population are not completely elucidated, our data suggest that clinical guidelines and educational interventions created by a multidisciplinary team can have a significant impact on antimicrobial use. Easily accessible antibiotic policy posters, smartphone friendly document of the antibiotic policy and frequent reminders to residents and faculty regarding usage of antibiotic policy contributed to the increase in compliance in this study.
In our study, the compliance improved in initial months and again dropped. Change in residents who decide on the initial choice of antibiotics, reduced reminders on antibiotics policy and difficulty in making specific diagnosis at the first point of contact with healthcare especially in case of pneumonia may be the reasons for drop in compliance. Thus, for sustaining compliance among residents we need to improve the skills of making a specific diagnosis and enhancing their confidence in the unit policy, continue regular daily monitoring of compliance and also revisiting the reasons for non-compliance and their prompt solutions. Fall in adherence in the end could also be due to seasonal influenza and other diseases which increase antibiotic prescriptions making adherence weak. The change in outcome might not be easily seen as antimicrobial resistance is multifactorial and rational use is just one of them.
There are several limitations to our study. First, the interventions were started in peak season for respiratory viral illnesses in September 2018 whereas the baseline data were collected from July 2017 to February 2018 and sustainability of intervention has been documented between October 2018 and March 2019. Also, there is only one point of comparison before the intervention and four points after intervention. Due to frequent rotation of the residents every 3 months, the impact on knowledge was not possible. Observation of the effects beyond 6 months would be beneficial in understanding the sustainability of the results of educational intervention.
Ours is a project involving educational intervention. With trainee residents getting rotated to different units every 3 months, we could not check the knowledge change but just measured the change in prescribing patterns.
Many of the senior paediatric consultants in our hospital were involved in the development of the antibiotic policy. The ‘buy-in’ effect of these key opinion leaders might have played an important role in our success. However, this also demonstrates that everyone should be involved in implementation of policies intended to be influenced by several factors and even subjective decisions.
We found that quality improvement interventions can lead to improved antibiotic prescribing practices of resident doctors for hospitalised children and it has no negative consequences. The use of specific antibiotics at an appropriate dose, frequency and duration is important for timely management and to prevent antimicrobial resistance. Although the medical professionals were willing to follow the antibiotic policy, it needed constant and persistent reinforcement and monitoring to keep the improvement sustainable. While the hunt is still on for the best methods for antimicrobial stewardship, the quality improvement initiative offers a plausible way in the right direction and can be easily reproduced in many other secondary and tertiary centres.
Data availability statement
Data are available on reasonable request. Analysed data are available in the manuscript.
Patient consent for publication
The study was approved by the institutional ethics committee (approval no: LHMC/ECHR/2017/96).
Contributors All authors were included in: conception and design of the work, data analysis and interpretation, drafting the article, critical revision of the article and inal approval of the version to be published. Kushala was involved in the above and with data collection as well. Dr Harish Pemde is responsible for the overall content as the guarantor for the article.
Funding Publication of this article is made open access with funding from the Nationwide Quality of Care Network.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.