Article Figures & Data
Figures
- Figure 1. Possible pathways for GP-level influence on diagnosis and prescription.
- Figure 2. Selection of patient for inclusion in the analysis. aToo few consultations per clinician. bImplausibly high numbers of consultations per clinician.
- Figure 3. Relationship between propensity to prescribe antibiotics and GP choice of diagnosis. Q = quintile.
Tables
Sore throat Pharyngitis or tonsillitis N (% of total) 232 652 (59.1) 160 938 (40.9) n % n % Antibiotics 89 649 38.5 134 811 83.8 Age band, years 0–4 18 086 7.8 44 632 27.7 5–9 24 069 10.3 26 379 16.4 10–19 33 133 14.2 24 783 15.4 20–29 27 188 11.7 19 222 11.9 30–39 36 341 15.6 21 058 13.1 40–49 30 003 12.9 11 775 7.3 50–59 24 675 10.6 6591 4.1 60–69 20 481 8.8 3984 2.5 70–79 12 534 5.4 1770 1.1 80–101 6142 2.6 744 0.5 Sex Male 95 855 41.2 72 968 45.3 Female 136 797 58.8 87 970 54.7 - Table 2. Random-effects models investigated for prediction of diagnosis (pharyngitis/tonsillitis or sore throat) and antibiotic prescribing
Models predicting diagnosis AIC P value Variance ofrandom effect OR for diagnosis(95% CI) % variance attributable to GP Marginal R2, % Conditional R2, % R2c – R2m, % 1a Null model for diagnosis diagnosis ~ 477462.9 NA NA NA NA NA 1b Null model + GP random-effect diagnosis ~ null model + (1|staffid) 441144.2 <0.001 1.16 NA 26.0 13.1 32.3 19.2 Models predicting antibiotic prescribing AIC P value Variance ofrandom effect OR for antibioticprescribing (95% CI) % varianceattributable to GP MarginalR2, % ConditionalR2, % R2c – R2m,% Null model for antibiotic abx ~ null model 515680.4 NA NA NA NA NA 2b Diagnosis-only model for antibiotic abx ~ diagnosis + null model 437269.3 <0.001 NA 2.24 (2.22 to 2.26) NA NA 2c GP random-effect only model for antibiotic abx ~ null model + (1|staffid) 496961.3 <0.001 0.59 NA 15 5.8 17.7 11.9 2d and GP random-intercept model for antibiotic abx ~ diagnosis + null model + (1|staffid) 417323.6 <0.001 0.81 2.48 (2.46 to 2.50) 20 26.3 38.6 12.3 2e and GP random-slope model for antibiotic abx ~ diagnosis + null model + (1 + diagnosis|staffid) 412008.4 <0.001 0.73 2.60 (2.55 to 2.64) 18 26.3 42.7 16.4 a Null model includes patient age (in age bands), patient sex, day of week, month, clinician type (for example, GP partner, salaried GP, locum GP, practice nurse), clinician sex. P-values from likelihood ratio test. Marginal R2 = variance explained by fixed effects. (Conditional R2 – Marginal R2) = variance explained by random effects. AIC = Akaika information criterion. abx = antibiotic prescribing. CI = confidence intervals. GP = clinician (includes a small number of practice nurse consultations). NA = not available. OR = odds ratio.
Supplementary Data
- TM_10.3399BJGPO.2023.0056_Supp.pdf -
Supplementary material is not copyedited or typeset, and is published as supplied by the author(s). The author(s) retain(s) responsibility for its accuracy.
In this issue
Medicalising diagnoses and treatment preferences: a retrospective cohort study of throat-related consultations in electronic primary care records
Jump to section
Keywords
More in this TOC Section
Related Articles
Cited By...
Intended for Healthcare Professionals
