- Research article
- Open Access
- Open Peer Review
Diagnostic labelling as determinant of antibiotic prescribing for acute respiratory tract episodes in general practice
© van Duijn et al; licensee BioMed Central Ltd. 2007
- Received: 17 March 2007
- Accepted: 20 September 2007
- Published: 20 September 2007
Next to other GP characteristics, diagnostic labelling (the proportion of acute respiratory tract (RT) episodes to be labelled as infections) probably contributes to a higher volume of antibiotic prescriptions for acute RT episodes. However, it is unknown whether there is an independent association between diagnostic labelling and the volume of prescribed antibiotics, or whether diagnostic labelling is associated with the number of presented acute RT episodes and consequently with the number of antibiotics prescribed per patient per year.
Data were used from the Second Dutch National Survey of General Practice (DNSGP-2) with 163 GPs from 85 Dutch practices, serving a population of 359,625 patients. Data over a 12 month period were analysed by means of multiple linear regression analysis. Main outcome measure was the volume of antibiotic prescriptions for acute RT episodes per 1,000 patients.
The incidence was 236.9 acute RT episodes/1,000 patients. GPs labelled about 70% of acute RT episodes as infections, and antibiotics were prescribed in 41% of all acute RT episodes. A higher incidence of acute RT episodes (beta 0.67), a stronger inclination to label episodes as infections (beta 0.24), a stronger endorsement of the need of antibiotics in case of white spots in the throat (beta 0.11) and being male (beta 0.11) were independent determinants of the prescribed volume of antibiotics for acute RT episodes, whereas diagnostic labelling was not correlated with the incidence of acute RT episodes.
Diagnostic labelling is a relevant factor in GPs' antibiotic prescribing independent from the incidence of acute RT episodes. Therefore, quality assurance programs and postgraduate courses should emphasise to use evidence based prognostic criteria (e.g. chronic respiratory co-morbidity and old age) as an indication to prescribe antibiotics in stead of single inflammation signs or diagnostic labels.
- Multiple Linear Regression Analysis
- White Spot
- Antibiotic Prescribe
- Quality Assurance Program
- Anatomical Therapeutic Chemical Classification
Most antibiotics are prescribed in primary care with acute respiratory tract (RT) infections being the main indication . However, there is insufficient evidence to warrant its use for most of these infections, while they are in general self-limiting . Even in a low prescribing country such as the Netherlands there is an over-prescribing of antibiotics; about 50% of the antibiotic prescriptions for acute RT episodes are not in accordance with Dutch national guidelines [3–5]. Considering costs, side-effects and the growing resistance to pathogens, it is important to rationalise antibiotic prescribing as much as possible . More insight into the determinants of prescribing antibiotics for acute RT episodes is therefore necessary to optimise medical education and feedback procedures .
Several GP characteristics have been shown to be associated with overall antibiotic prescribing rates for acute RT infections (e.g. the number of years of practice, number of patients, perceived workload, and the perception that purulent sputum is an indication for antibiotic treatment) [8–11]. Next to this, there are indications that both diagnostic labelling (i.e. the tendency to encode acute RT episodes in medical records more as infections than as symptoms) and incidence of acute RT episodes (the number of acute RT episodes per 1,000 patients presented to the GP) are associated with the volume of antibiotic prescribing [12–16].
However, it is unknown whether there is an independent association between diagnostic labelling and the volume of prescribed antibiotics, or whether diagnostic labelling is associated with the number of presented acute RT episodes and therefore with the volume of prescribed antibiotics. After all, labelling acute RT episodes as diagnoses instead of symptoms might be a trigger for patients to revisit the GP on a subsequent occasion, i.e. GPs who are inclined to label acute RT episodes as infections might consequently be visited more frequently by patients for these episodes.
Studies that include both diagnostic labelling and the incidence of acute RT episodes as possible determinants of antibiotic prescribing combined with GP personal characteristics are lacking. Furthermore, there are hardly any studies about possible determinants of antibiotic prescribing using data from nation-wide electronic GP data bases. Therefore the present study explores if diagnostic labelling, the incidence of acute RT episodes presented to the GP, and other GP characteristics are associated with the volume of antibiotic prescribing for acute RT episodes. This study is based on data of a nationwide study .
GPs, practices and patients
The data used in the present study were derived from the Second Dutch National Survey of General Practice (DNSGP-2), carried out by the Netherlands Institute for Health Services Research (NIVEL) in 2001 . Data were used from 163 GPs from 85 practices serving a population of 359,625 patients. The patients enlisted in the participating practices were similar to the profile of the Dutch general population with respect to age, gender and type of health care insurance. There were no differences between the total population of Dutch GPs and the study group except for the type of practice: i.e. single-handed GPs where underrepresented in the study population. In all, the DNSGP-2 is assumed to provide a representative impression of the morbidity and prescribing habits in Dutch general practice.
Morbidity and prescribing
Acute respiratory tract episodes divided into localisation and type (with the corresponding ICPC codes)
Labelled as symptom:
Labelled as infection:
Upper respiratory tract:
Upper respiratory tract:
R07 sneezing/nasal congestion
R72 strep throat
R09 sinus symptom/complaint
R74 acute upper respiratory infection
R21 throat symptom/complaint
R75 sinusitis acute/chronic
R22 tonsils symptom/complaint
R76 acute tonsillitis
H01 ear pain/earache
H71 acute otitis media/myringitis
Lower respiratory tract:
Lower respiratory tract:
R77 acute laryngitis/tracheitis
In 22 of the 85 practices, because the prescriptions could not be linked to a specific GP within the practice, the average practice prescription rates were allocated to all GPs of that practice.
At the beginning of the DNSGP-2 information was collected about the GPs and their practices, including age (years), gender (male/female), years since registration as a GP (years), number of enlisted patients and type of practice (single-handed yes/no). Additional information was collected by means of a written questionnaire: frequency of consulting national GP guidelines (once a week or less/more than once a week), seeing pharmaceutical representatives in the four weeks preceding completion of the questionnaire (no/yes), inclination to prescribe new drugs ranging from 1 (low) to 5 (high), views on RT symptoms and antibiotics rated on a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree) [20, 21] and GPs' medical knowledge on acute RT infections and antibiotics (using a 10-question questionnaire; scored from 0 (very low) to 10 (very high)).
Outcome measure and analysis
The outcome measure was the prescribed volume of antibiotics (the number of antibiotic prescriptions per 1,000 patients per GP per year) for acute RT episodes. To explore the association between GP characteristics and the volume of antibiotic prescribing for acute RT episodes a multiple linear regression analysis was carried out after checking for interactions and collinearity. All determinants that had a bivariate correlation with the outcome measures at p < 0.20, were included in the multiple linear regression analysis with a stepwise procedure, followed by an enter procedure. The strength of the associations between the determinants and the volume of antibiotics was described by standardised beta coefficients with 95% confidential intervals (95% CI).
All data were analysed using the Statistical Package for Social Sciences for Windows (SPSS 12.0.1).
Data on the 163 general practitioners participating in the study
Age in years (mean (SD))
Gender (% female)
Years since registration (mean (SD))
Number of patients (mean (SD))
Type of practice (% single-handed)
Consulting national guidelines for GPs (% > once a week)
Seeing pharmaceutical representatives (% yes)
Inclination to prescribe new drugs (mean (SD))*
Medical knowledge on respiratory tract symptoms and antibiotics (mean (SD))**
Views on acute respiratory tract symptoms and antibiotics (mean (SD))***
- Self- limiting character
- Need to consult a general practitioner
- Need of antibiotics in case of fever
- Need of antibiotics in case of green phlegm
- Need of antibiotics in case of white spots in the throat
- Effectiveness of antibiotics
- Side-effects of antibiotics
Acute respiratory tract episodes/1,000 patients/year (mean (SD))
Proportion of acute respiratory tract episodes labelled as diagnosis (mean (SD))
Antibiotic prescriptions for acute respiratory tract episodes per 1,000 patients/year (mean (SD))
Proportion of antibiotic prescriptions/acute respiratory tract episodes (mean (SD))
Morbidity and prescribing
In total 236.9 acute RT episodes/1,000 patients were registered (Table 2). More than half of these episodes were for upper acute RT episodes and the remainder was for lower acute RT episodes (150.3 vs. 86.6 episodes/1,000 patients). GPs labelled 70% of all acute RT episodes as infections. In 41% of all acute RT episodes antibiotics were prescribed (97.3 antibiotic prescriptions related to 236.9 acute RT episodes), with no differences between lower and upper acute RT episodes.
Determinants of antibiotic prescribing
Correlation between GPs' characteristics and volume of antibiotic prescriptions per 1,000 patients for acute respiratory tract episodes*
Variables in model
Standardised beta coefficient
Acute respiratory tract episodes/1,000 patients/year
Proportion of acute respiratory tract episodes labelled as infection
GP's endorsement of the need of antibiotics in case of white spots in the throat
Variance explained by the model (R 2 )
Strengths and weaknesses of the study
The 163 participating GPs were representative for the total population of GPs in the Netherlands. The GPs' characteristics were comparable with those of the total population of Dutch GPs, except for the type of practice; single-handed GPs were under-represented in the present study. However, because this factor was not significantly correlated with the volume of antibiotics prescribed for acute RT episodes per 1,000 patients, it is unlikely that this influenced the results. In addition, the patients involved in the DNSGP-2 study reflected the general Dutch population .
We assume that the morbidity and antibiotic prescribing data are accurate because they were extracted from the electronic medical records of the participating practices, and the inter-observer reliability of coding episodes into the ICPC codes is high . The registration covered a 12-month period for each practice, thereby eliminating seasonal influences. Considering the representativeness of the participating GPs and their patients – and the high validity of the data – the results of the present study can be assumed to validly represent morbidity and GPs prescribing behaviour in Dutch general practice.
For 56 of the 163 GPs (in 22 of the 85 practices) the volume of antibiotics prescribed was only available on the practice level. In these cases the volume of antibiotics prescribed by each GP was estimated by means of the number of prescriptions per practice. This implies a loss of variance in outcome measures so that associations in the regression analysis might have been underestimated. However, analysis of 107 GPs (163 minus 56) yielded comparable results. Finally, because our study had a cross-sectional design, we can only assume correlations and no conclusions about causal relationships can be drawn.
Comparison with existing literature
In the present study 70% of all acute RT episodes are labelled as infections, compared with 41%–62% in other studies that labelled their acute RT episodes with a diagnosis 'assuming a bacterial infection' [13, 14]. In a recent Dutch study 63% of the acute RT episodes were labelled as infections, and the antibiotic prescribing rate for all acute RT episodes was about 35% . Compared with the latter study in which data were collected by hand-written prospective recording of visits we assume that our data collection from the electronic medical records is probably more accurate.
Our finding that the number of acute RT episodes presented to the GP/1,000 patients is strongly associated with the volume of prescribed antibiotics/1,000 patients for acute RT episodes has also been suggested by Ashworth et al . Diagnostic labelling also turns out to be an independent determinant of the volume of antibiotics prescribed for acute RT episodes; this latter phenomenon has already been described by Howie in 1983 , and later confirmed by others [11, 13–15]. With a correlation between the incidence of acute RT episodes and diagnostic labelling being absent we point that diagnostic labelling is directly related with the volume of antibiotic prescriptions with the number of presented acute RT episodes not being an intermediate factor.
Evidently, while there are no geographical, biological or medical explanations of variation in labelling acute RT episodes as infections in the Netherlands, we propose that diagnostic labelling is an arbitrary process, undoubtedly partially used to justify antibiotic prescribing. GPs may use this mechanism to defend themselves against unforeseen complications or worsening, even though these sequels will very seldom occur .
Signs of inflammation (such as white spots in the throat) was also found to be a determinant for over-prescribing of antibiotics in another Dutch study . This supports the importance of a more selective indication setting in antibiotic prescribing. After all, as stated above, acute respiratory tract infections seldom need antibiotics, even if these infections are accompanied by green phlegm or white spots in the throat .
The more acute RT episodes/1,000 patients are presented to a GP and the more GPs label these episodes as infections rather than as symptoms, the more GPs prescribe antibiotics for acute RT episodes, while being a male GP and endorsing the need of antibiotics in case of white spots in the throat are predictors too. This implicates that diagnostic labelling is a relevant factor in GPs' antibiotic prescribing independent from the incidence of acute RT episodes. Therefore, quality assurance programs and postgraduate courses should emphasise to use evidence based prognostic criteria (e.g. chronic respiratory co-morbidity and old age) as an indication to prescribe antibiotics in stead of single inflammation signs or diagnostic labels. Structured peer review groups combined with (postgraduate) education of GPs may be suitable methods to implement such recommendations.
At last, since diagnostic labelling might be influenced by e.g. reasons related to therapeutic justification, the use of diagnostic labels in research has to be handled with some caution.
- Kuyvenhoven MM, van Balen FAM, Verheij TJM: Outpatient antibiotic prescriptions from 1992 to 2001 in the Netherlands. J Antimicrob Chemother. 2003, 52: 675-8. 10.1093/jac/dkg412.View ArticlePubMedGoogle Scholar
- Fahey T, Howie J: Re-evaluation of a randomized controlled trial of antibiotics for minor respiratory illness in general practice. Fam Pract. 2001, 18: 246-8. 10.1093/fampra/18.3.246.View ArticlePubMedGoogle Scholar
- Coenen S, Kuyvenhoven MM, Butler CC, van Royen P, Verheij TJM: Variation in European antibiotic use. Lancet. 2001, 358: 1272-10.1016/S0140-6736(01)06364-4.View ArticlePubMedGoogle Scholar
- Akkerman AE, Kuyvenhoven MM, van der Wouden JC, Verheij TJM: Analysis of under- and overprescribing of antibiotics in acute otitis media in general practice. J Antimicrob Chemother. 2005, 56: 569-74. 10.1093/jac/dki257.View ArticlePubMedGoogle Scholar
- Akkerman AE, Kuyvenhoven MM, van der Wouden JC, Verheij TJM: Determinants of antibiotic overprescribing in respiratory tract infections general practice. J Antimicrob Chemother. 2005, 56: 930-6. 10.1093/jac/dki283.View ArticlePubMedGoogle Scholar
- Goossens H, Ferech M, vander Stichele R, Elseviers M: Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet. 2005, 365: 579-87.View ArticlePubMedGoogle Scholar
- Welschen I, Kuyvenhoven MM, Hoes AW, Verheij TJM: Effectiveness of a multiple intervention to reduce antibiotic prescribing for respiratory tract symptoms in primary care: randomised controlled trial. BMJ. 2004, 329: 431-5. 10.1136/bmj.38182.591238.EB.View ArticlePubMedPubMed CentralGoogle Scholar
- Kuyvenhoven MM, de Melker RA, van der Velden K: Prescription of antibiotics and prescribers' characteristics. A study in prescription of antibiotics in upper respiratory tract infections in general practice. Fam Pract. 1993, 10: 366-7. 10.1093/fampra/10.4.366.View ArticlePubMedGoogle Scholar
- Gonzales R, Barrett PH, Steiner JF: The relation between purulent manifestations and antibiotic treatment of upper respiratory tract infections. J Gen Int Med. 1999, 14: 151-6. 10.1046/j.1525-1497.1999.00306.x.View ArticleGoogle Scholar
- Murray S, Del Mar C, O'Rourke P: Predictors of an antibiotic prescription by GPs for respiratory tract infections: a pilot. Fam Pract. 2000, 17: 386-88. 10.1093/fampra/17.5.386.View ArticlePubMedGoogle Scholar
- Westert GP, Schellevis FG, De Bakker DH, Groenewegen PP, Bensing JM, van der Zee J: Monitoring health inequalities through general practice: the Second Dutch National Survey of General Practice. Eur J Public Health. 2005, 15: 59-65. 10.1093/eurpub/cki116.View ArticlePubMedGoogle Scholar
- Howie JGR, Richardson IM, Gill G, Durno D: Respiratory illness and antibiotic use in general practice. J R Coll Gen Pract. 1971, 21: 657-63.PubMedPubMed CentralGoogle Scholar
- Cars H, Hakansson A: To prescribe – or not to prescribe-antibiotics. District physicians' habits vary greatly, and are difficult to change. Scand J Prim Health Care. 1995, 13: 3-7.View ArticlePubMedGoogle Scholar
- Hutchinson JM, Jelinski S, Hefferton D, Desaulniers G, Parfrey PS: Role of diagnostic labeling in antibiotic prescription. Can Fam Physician. 2001, 47: 1217-24.PubMedPubMed CentralGoogle Scholar
- De Sutter AI, De Meyere MJ, De Maeseneer JM, Peersman WP: Antibiotic prescribing in acute infections of the nose or sinuses: a matter of personal habit?. Fam Pract. 2001, 18: 209-13. 10.1093/fampra/18.2.209.View ArticlePubMedGoogle Scholar
- Kuyvenhoven MM, Verheij TJM, de Melker RA, van der Velden J: Antimicrobial agents in lower respiratory tract infections in Dutch general practice. Br J Gen Pract. 2000, 50: 133-4.PubMedPubMed CentralGoogle Scholar
- Ashworth M, Cox K, Latinovic R, Charlton J, Gulliford M, Rowlands G: Why has prescribing for respiratory illness declined in primary care? A longitudinal study using the General Practice Research Database. J Public Health. 2004, 26: 268-74. 10.1093/pubmed/fdh160.View ArticleGoogle Scholar
- Lamberts H, Woods M, (eds): The International Classification of Primary Care. 1987, London: Oxford University PressGoogle Scholar
- WHO Collaborating Centre for Drug Statistics Methodology: Anatomical Therapeutic Chemical classification. [http://www.whocc.no/atcddd/]
- van Duijn H, Kuyvenhoven M, Welschen I, den Ouden H, Slootweg A, Verheij T: Patients' and doctors' views on respiratory tract symptoms. Scand J Prim Health Care. 2002, 20: 201-2. 10.1080/028134302321004836.View ArticlePubMedGoogle Scholar
- van Duijn HJ, Kuyvenhoven MM, Schellevis FG, Verheij TJM: Views on respiratory tract symptoms and antibiotics of Dutch general practitioners, practice staff and patients. Patient Educ Couns. 2006, 61: 342-7. 10.1016/j.pec.2005.03.012.View ArticlePubMedGoogle Scholar
- van der Linden MW, Westert GP, de Bakker DH, Schellevis FG: Tweede nationale studie naar ziekten en verrichtingen in de huisartsenpraktijk: klachten en aandoeningen in de bevolking en in de huisartspraktijk. [Second Dutch national survey on morbidity and interventions in general practice: complaints and diseases in the population and in general practice; in Dutch]. 2004, Utrecht/Bilthoven: NIVEL/RIVMGoogle Scholar
- Akkerman AE, Kuyvenhoven MM, van der Wouden JC, Verheij TJM: Prescribing antibiotics for respiratory tract infections by GPs: management and prescriber characteristics. Br J Gen Pract. 2005, 55: 114-8.PubMedPubMed CentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://0-www.biomedcentral.com.brum.beds.ac.uk/1471-2296/8/55/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.