Benchmarking of Viral Bronchiolitis Management by General Practitioners in the United Kingdom

Abstract

Viral bronchiolitis is the leading cause of hospitalization in infants in the United Kingdom (UK) with wide variation in rates of hospitalization in different geographical regions of the UK. A potential cause of these differences is variation in primary care management and referral to hospital. This study aimed to prospectively survey general practitioners (GPs) in the UK to provide a benchmark of practice against which future practice can be assessed. An electronic, structured questionnaire was sent to 1,001 geographically representative GPs in primary care centers in the UK, through the market research company MedeConnect, to assess their management of infants with viral bronchiolitis. We measured practice before the 2015 National Institute for Health and Care Excellence (NICE) bronchiolitis guideline against the guideline, to obtain a benchmark of practice. We also used a multivariate analysis to assess GP factors associated with variation in management. Thirty-nine percent of GPs did not refer to any guideline to manage infants with bronchiolitis, 33% did not routinely measure oxygen saturations, 48% prescribed an “inappropriate” (evidence of no benefit) medication, and 62% did not give written guidance to parents. GP factors influencing management included the year the GP qualified, sex, region of practice, and working at a dispensing practice. Up to 75% of GPs' management did not conform to the newly published 2015 NICE bronchiolitis guideline before its publication. There was wide variation in the management of infants with viral bronchiolitis by UK GPs. Most infants with viral bronchiolitis are not managed optimally by GPs and multiple GP factors influenced this management.

Introduction

Viral bronchiolitis, especially that caused by respiratory syncytial virus (RSV), is the leading cause of hospitalization in infants under a year old. More than 30,000 infants are hospitalized yearly in England and Wales due to viral bronchiolitis with a wide variation in rates of hospitalization in different United Kingdom (UK) geographical regions.^1–3 No clear cause has been identified for this variation although potential factors include socioeconomic deprivation and differences in primary care referral and hospital admission guidelines.²

Management of bronchiolitis is purely supportive (oxygen supplementation and feeding support). There is evidence that medications, including bronchodilators, steroids, leukotriene receptor antagonists, and antibiotics,^4–7 have no role in the vast majority of bronchiolitic infants. There are currently no vaccines for the viruses causing bronchiolitis, although several against RSV are in development.⁸ Passive immunization with a monoclonal antibody (palivizumab) is available for high-risk infants.

The Scottish Intercollegiate Guideline Network (SIGN) published a guideline for managing infants with bronchiolitis in primary and secondary care in 2006,⁹ but until June 1, 2015 when the National Institute for Health and Care Excellence (NICE) published their guidance¹⁰ there was no national guidance in England and Wales. The guidance given by NICE and SIGN is very similar, with the main difference being the level of oxygen saturation requiring referral to hospital from primary care; <95% in SIGN and <92% in NICE. Individual general practitioner (GP) surgeries have local guidelines often based on national (SIGN) or international [American Academy of Pediatrics (AAP)] guidelines, but how variable these guidelines are in the United Kingdom (UK) National Health Service (NHS) is unknown.

The aim of this study was to prospectively survey GPs in the UK (England, Scotland, Wales, and Northern Ireland) to provide a benchmark of practice before the publication of the 2015 NICE bronchiolitis guideline against which future practice can be assessed. We also aimed to investigate possible GP causes of any variation in management.

Materials and Methods

Over a 2-week period in March 2015, an electronic, structured questionnaire based on the NICE bronchiolitis draft guideline, was sent to GPs registered with the General Medical Council in the UK and recruited by MedeConnect, the market research division of Doctors.net.uk. A sample of 1,000 GPs stratified by UK geographic region was requested. MedeConnect sent invitations to members followed by reminders until 1,000 GPs had completed the questionnaire. GPs were not aware of the topic of the questionnaire, so it was not biased by being completed by GPs with an interest in pediatrics. Demographic data were collected from GPs and five research questions were included to investigate the management of infants with bronchiolitis (Supplementary Box S1; Supplementary materials are available online at www.liebertpub.com/ped).

We compared the proportions of GPs from individual countries (England, Wales, Scotland, and Northern Ireland) who would be compliant with the NICE guideline.¹⁰ The NICE guideline¹⁰ suggests all infants should be referred to hospital if oxygen saturations are <92% in air (if GPs did not assess oxygen saturations they were deemed noncompliant with the guideline) or if feeding is <50–75% of normal (we considered feeding <50% of normal an absolute referral criteria), antibiotics, inhaled bronchodilators, leukotriene receptor antagonists, and steroids (oral/inhaled) should not be prescribed and parents should be provided with “key safety information to take away for reference” (we interpreted that as needing to provide written advice).

Statistical analysis

Proportions were compared using the chi-squared test or Fisher's exact test as appropriate. Post hoc testing was carried out using the Kruskal–Wallis test of significance. A multivariate analysis, by means of a logistic regression model, was carried out on the dichotomized responses against the categorical covariates. Correlation coefficients were calculated to assess whether hospital admission rates in English NHS Strategic Health Authorities were influenced by various GP management factors. Statistical analysis was carried out with IBM SPSS Statistics (version 22, New York).

This study did not require ethical approval.

Results

In total 1,001 GPs completed the questionnaire, 552 (55%) were male. Participants included 629 (63%) GP principals, 252 (25%) salaried GPs, 119 (12%) locum GPs, and one (0.1%) trainee GP. Eighty-nine (9%) qualified before 1980, 219 (22%) between 1980 and 1989, 330 (33%) between 1990 and 1999 and 363 (36%) between 2000 and 2010. The median (range) number of GPs in each practice was six (1–30) and the median (range) GP practice size was 8,100 (500–36,000) patients. One hundred eighty-seven (19%) GP practices were dispensing practices (ie, had a pharmacy on site). The geographic areas where GPs practiced are shown in Supplementary Table S1; 399 (40%) were in urban areas, 254 (25%) suburban areas, 238 (24%) semirural areas, 102 (10%) rural areas, and eight (1%) defined as “other” areas (eg, included prisons and university campuses).

Two hundred ninety (29%) GPs used a locally developed guideline, 248 (25%) used a commercially available guideline (eg, from an internet website, www.gpnotebook.co.uk), 227 (23%) used a national guideline (ie, SIGN), and 390 (39%) did not use any guideline. Some GPs used more than one guideline. Supplementary Figure S1a shows the geographic variation in use of a guideline by GPs.

If there were no other indications for referral to hospital, four (0.4%) GPs routinely referred bronchiolitic infants to hospital only if their oxygen saturations were ≤88%, 60 (6%) at ≤90%, 250 (25%) at ≤92%, and 360 (36%) at ≤94%. Three hundred twenty-seven (33%) GPs did not test oxygen saturations on bronchiolitic infants. Supplementary Figure S1b and c show the geographic variation in use of oxygen saturations by GPs and in guidelines for referral to hospital when oxygen saturations are ≤92% (including those who only refer if saturations are ≤92%, ≤90%, or ≤88%).

If there were no other indications for referral to hospital, 145 (14%) GPs routinely referred bronchiolitic infants to hospital only if they were feeding <33% of normal, 564 (56%) if feeding <50% of normal, 190 (19%) if feeding <75% of normal, and 14 (1%) if feeding was anything less than normal. Supplementary Figure S1d shows the geographic variation in referral to hospital if feeding <50% of normal.

Three hundred forty-nine (35%) GPs routinely prescribed inhaled bronchodilators for bronchiolitic infants, 53 (5%) inhaled steroids, 195 (19%) oral steroids, 75 (7%) oral antibiotics, 16 (2%) oral leukotriene receptor antagonists, 224 (24%) saline (or similar) nose drops, and 435 (43%) prescribed no medicines routinely. Supplementary Figure S1e shows the geographic variation in use of inappropriate medications (any of the above medications, excluding saline nasal drops, which should not be prescribed according to the NICE guideline).

One hundred twenty-four (12%) GPs routinely provided locally developed written guidance for parents of bronchiolitic infants, 99 (10%) nationally developed guidance (eg, the SIGN information leaflet), 156 (16%) provided written guidance from another source, and 622 (62%) did not routinely provide written guidance. Supplementary Figure S1f shows the geographic variation in provision of written advice.

We compared the geographic variation with each of the above factors to the hospitalization rate for bronchiolitis by NHS strategic health authority calculated in a study by Murray et al.³ (Supplementary Fig. S2). None of the above factors explained the variation in hospitalization rate (all P values >0.05, Supplementary Fig. S1a–e).

Only 58% of GPs in Scotland, 62% in England, 57% in Wales, and 44% in NI (P = 0.25) referred to any guideline for the management of infants with bronchiolitis. In Scotland 23% used the SIGN guideline compared with 23% in England, 24% in Wales, and 22% in NI (P = 0.99). Table 1 shows the current compliance with various aspects of the NICE guideline by the country in which the GP works. There were significant differences in the percentage of GPs prescribing inhaled bronchodilators (P = 0.01) and providing written advice (P = 0.03) between the four countries.

Table 1.

Percentage of General Practitioners Compliant with the National Institute for Health and Care Excellence Guideline by Country

	Total (%)	England (%)	Scotland (%)	Wales (%)	NI (%)	P^a
Hospital referral if oxygen saturations ≤92% (excluding those who only referred once saturations ≤90% or ≤88%)^b	25	26	17	14	30	0.40
Hospital referral if feeding <50% (excluding those who only referred once feeding <33%)	56	55	62	57	63	0.18
Use of antibiotics	93	93	94	91	81	0.10
Use of inhaled bronchodilator	65	64	78	67	48	0.01
Use of oral steroid	81	80	80	84	78	0.95
Use of inhaled steroid	95	95	95	94	89	0.58
Use of montelukast	98	98	99	100	96	0.70
Providing an advice sheet	38	40	27	29	44	0.03

The P values compare proportions across the four countries.

If using the Scottish Intercollegiate Guideline Network guideline (referral to hospital if oxygen saturations <95%) then the compliance was 34% England, 46% Scotland, 39% Wales, 41% Northern Ireland, and 36% overall, P = 0.13.

Factors influencing management

Multivariate logistic analysis was carried out on each dichotomous response to determine which covariates had the largest influence on the response when accounting for the influence of other covariates (Supplementary Table S2). The covariate which influenced most of the responses was the year in which a GP qualified. This covariate was significant in predicting whether a guideline was used, whether an infant was referred to hospital if feeding was <50%, whether oral steroids or antibiotics were prescribed, whether any inappropriate medication was prescribed, and whether written advice was given.

The region of the UK where the GP practice was located had a significant influence on the use of oxygen saturation monitoring, on whether salbutamol and saline nasal drops were prescribed, and on whether any inappropriate medication was prescribed.

Sex of the GP was significantly associated with whether an infant was referred if the oxygen saturations were <92% and if feeding was <50%. Male GPs were more than twice as likely as females to refer infants to hospital only once the oxygen saturations were <92% (odds ratio [OR] confidence interval [95% CI] 2.44 [1.66–3.32]), and one and a half times more likely to refer infants to hospital only once the feeding was <50% of normal (OR 1.48 [1.10–2.01]). As the number of GPs in the practice increased, the likelihood of referring the infant to hospital only once their oxygen saturations were <92% increased (OR 1.62 [1.18–2.22]) and the odds of prescribing salbutamol decreased (OR 0.74 [0.57–0.97]).

Working in a dispensing practice was significantly associated with the prescription of oral leukotriene receptor antagonists (eg, montelukast) (OR 11.63 [2.62–52.63]). The type of GP location was associated with whether the GP referred the infant to hospital only once feeding was <50% of normal. GPs in rural regions were more likely to refer infants if feeding was still >50% of normal. Compared with GPs working in rural locations, GPs in urban areas were almost twice as likely to wait until feeding was <50% before referring infants, although this was only weakly significant (OR 1.93 [0.99–3.75]). GPs in semiurban areas were more than twice as likely to refer infants only once feeding was <50% of normal than GPs in rural locations (OR 2.34 [1.26–4.42]).

Discussion

This study has demonstrated a wide variation in the guidelines and management of infants with bronchiolitis in primary care in the UK. Thirty-nine percent of GPs did not use any guideline to manage bronchiolitis, and of those that did the source of the guideline was variable. The NICE guideline aims to reduce this variation.¹⁰ However, despite the similar SIGN guideline⁹ having been available in Scotland (and also used in other regions of the UK) since 2006, management of bronchiolitis still varied considerably from that advised.

Few studies have addressed the impact of a clinical guideline on the management of bronchiolitis in primary care. One study¹¹ in France demonstrated the introduction of a clinical guideline resulted in a small but significant reduction in the prescription of steroids, bronchodilators, and mucolytics and a statistically significant 28% absolute increase (from 29% to 57%) in the provision of “adequate general advice.” There was no significant impact on the use of investigations.¹¹

In our study, use of inappropriate medications ranged from 2% to 35% depending on the medication. We found 19% of GPs routinely prescribed oral steroids. This is concerning as there is good evidence oral steroids are of no benefit to infants with bronchiolitis⁶ (or indeed in most circumstances in the main differential diagnosis of virus-induced wheeze¹²), and steroids potentially have a negative impact on growth and bone mineral density.¹³ Only 7% of GPs, however, routinely prescribed antibiotics.

In our study, factors that were demonstrated in multivariate analysis to be related to variation in management were the sex and year of qualification of the GP, whether the GP practice was a dispensing practice, the location and type of location of the GP practice, and the number of GPs working in the practice. More experienced GPs were less likely to use a guideline, more likely to prescribe inappropriate medications (oral steroids and antibiotics) and had a lower threshold for referral to hospital. This suggests more experienced GPs are either unfamiliar with newer evidence and clinical guidelines, are more confident in managing these patients but less risk averse, prescribe more for potential differential diagnoses, or give a prescription which is not clinically indicated.

One study,¹⁴ before the publication of the AAP bronchiolitis guideline in 2006, found that, although more experienced U.S. pediatricians were less risk averse than more junior colleagues, how risk averse a pediatrician was had no impact on their rate of admitting infants with bronchiolitis.

We have also demonstrated that different geographic areas were associated with marked variation in the primary care management of bronchiolitis. When comparing these factors with hospital admission rates by the same geographic areas from the study by Murray et al.,³ no clear pattern was identified. This suggests that these factors individually are not associated with the geographic variation in hospitalization rates, the geographic areas studied are too large to note local differences or other factors not easily measurable, for example, the overall subjective clinical and social picture of the child and family, may have a role irrespective of objective clinical parameters.

GPs working in dispensing practices were more likely to prescribe leukotriene receptor antagonists and those working in practices with fewer GPs were more likely to prescribe bronchodilators. A previous study,¹⁵ which surveyed UK GPs, found those working in dispensing practices, male GPs, GPs working in deprived areas, single-handed GPs, GP trainers, and GPs with increased length of service, had higher prescribing costs, similar to our findings for bronchiolitis.

The explanation for these variations was multifactorial but could have included knowledge of alternative and newer medications, incentives to increase profitability through prescribing in dispensing practices, and time pressure leading to “giving a prescription which is not clinically indicated.”¹⁵

The less time a GP has with the patient, the less carefully they are able to assess clinical parameters, the less time they spend explaining the diagnosis and possible outcomes to the family, and the more likely they are to prescribe an inappropriate medication to expedite the completion of the consultation. Inappropriate prescriptions add to healthcare costs and can worsen outcome (eg, the potential for worsening oxygen saturations if an infant struggles against an inhaler mask pressed on the face).

In addition, a poor explanation of the condition to the family may result in increased costs as either parents may seek further, unnecessary healthcare reviews or not reattend soon enough if their child deteriorates and therefore require more intensive management.

This study has a number of strengths. We have undertaken the largest survey of UK GPs investigating the management of bronchiolitis. We surveyed over 1,000 GPs with a representative geographic distribution across the UK. We included GPs with a wide range of experiences in a variety of settings and thus were able to assess the relationship various GP factors had on the management strategies used.

The study also had several weaknesses. The GPs were recruited through Doctors.net.uk and may not necessarily be representative of all GPs. In particular, we did not include trainee GPs. This methodology, however, has been used previously to survey GPs in the UK.^16,17 We did not look at individual patient notes and thus cannot comment on whether what GPs reported they did actually manifest in patient care. We may have, therefore, over- or underestimated guidance suggesting the use of medications and investigations.

In conclusion, we have shown wide variation in the guidelines and management of infants with bronchiolitis in primary care in the UK and provided a benchmark against which future studies can be assessed. Many GPs were not compliant with the new NICE guideline before its publication meaning they need to update their guidelines and practices. We welcome the new NICE guideline, however, ensuring it is widely implemented and adhered to will be challenging. As has been shown in the United States with the AAP bronchiolitis guideline,¹⁸ publishing a national guideline does not necessarily result in high compliance with it. We plan to carry out repeat surveys of GPs to assess the impact of the NICE guideline in its first years after implementation.

Footnotes

Acknowledgments

This study was funded by the Medical Sciences Division Medical Research Fund and the Jenner Institute of the University of Oxford. The funders had no role in the design, data collection or analysis, interpretation, write up or decision for article submission of the project. The authors wish to thank the general practitioners who completed the survey and MedeConnect for collecting data.

Contributorship

S.B.D. and A.J.P. designed the study. S.B.D. and E.P.G. devised the surveys. S.B.D. and A.N. analyzed the data. S.B.D. wrote the initial draft and all authors revised the draft.

Author Disclosure Statement

A.J.P. has previously conducted vaccine clinical trials on behalf of Oxford University funded by vaccine manufacturers, but he no longer does so and did not receive any personal reimbursement from them. A.J.P. is chair of the Department of Health's (DH) Joint Committee on Vaccination and Immunization (JCVI), but the reviews expressed herein do not necessarily represent those of DH or JCVI. S.B.D., A.N. and E.P.G. have no conflicts of interest to declare.

References

Green

, Yeates

, Goldacre

, et al. Admission to hospital for bronchiolitis in England: trends over five decades, geographical variation and association with perinatal characteristics and subsequent asthma. Arch Dis Child, 2016; 101:140–146.

Cheung

, Smith

, Thurland

, et al. Population variation in admission rates and duration of inpatient stay for bronchiolitis in England. Arch Dis Child, 2013; 98:57–59.

Murray

, Bottle

, Sharland

, et al. Risk factors for hospital admission with RSV bronchiolitis in England: a population-based birth cohort study. PLoS One, 2014; 9:e89186.

Gadomski

, Scribani

. Bronchodilators for bronchiolitis. Cochrane Database Syst Rev, 2014; 6:CD001266.

Farley

, Gkp

, Eriksson

, Cb

. Antibiotics for bronchiolitis in children under two years of age. Cochrane Database Syst Rev, 2014; 10:CD005189.

Fernandes

, Bialy

, Vandermeer

, et al. Glucocorticoids for acute viral bronchiolitis in infants and young children. Cochrane Database Syst Rev, 2013; (6):CD004878.

Liu

, Ouyang

, Sharma

, et al. Leukotriene inhibitors for bronchiolitis in infants and young children. Cochrane Database Syst Rev, 2015; 3:CD010636.

Murray

, Saxena

, Sharland

. Preventing severe respiratory syncytial virus disease: passive, active immunisation and new antivirals. Arch Dis Child, 2014; 99:469–473.

Scottish Intercollegiate Guidelines Network (SIGN). Bronchiolitis in children. SIGN guideline 91 (2006). ISBN 1 (10) 905813 01 5.

10.

National Institute for Health and Care Excellence. Bronchiolitis: diagnosis and management of bronchiolitis in children. Clinical Guideline NG 9. 2015.

11.

Touzet

, Réfabert

, Letrilliart

, et al. Impact of consensus development conference guidelines on primary care of bronchiolitis: are national guidelines being followed?. J Eval Clin Pract, 2007; 13:651–656.

12.

Bush

, Grigg

, Saglani

. Managing wheeze in preschool children. BMJ, 2014; 348:g15.

13.

Guilbert

, Morgan

, Zeiger

, et al. Long-term inhaled corticosteroids in preschool children at high risk for asthma. N Engl J Med, 2006; 354:1985–1997.

14.

Baldwin

, Green

, Shaw

, et al. Physician risk attitudes and hospitalization of infants with bronchiolitis. Acad Emerg Med, 2005; 12:142–146.

15.

Watkins

, Harvey

, Carthy

, et al. Attitudes and behaviour of general practitioners and their prescribing costs: a national cross sectional survey. Qual Saf Health Care, 2003; 12:29–34.

16.

Plüddemann

, Wallace

, Bankhead

, et al. Clinical prediction rules in practice: review of clinical guidelines and survey of GPs. Br J Gen Pract, 2014; 64:233–242.

17.

Ahmad

, Orrell

, Iliffe

, et al. GPs' attitudes, awareness, and practice regarding early diagnosis of dementia. Br J Gen Pract, 2010; 60:360–365.

18.

Parikh

, Hall

, Teach

. Bronchiolitis management before and after the AAP guidelines. Pediatrics, 2014; 133:e1–e7.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.13 MB

0.04 MB

0.03 MB