Abstract / Description of output
Background
Variation in general practice (GP) referral rates to outpatient services is well described however variance in rates of referral to acute medical units is lacking.
Objective
To investigate variance in GP referral rate for acute medical assessment and subsequent need for hospital admission.
Methods
A retrospective cohort study of acute medical referrals from 88 GPs in Lothian, Scotland between 2017 and 2020 was performed using practice population size, age, deprivation, care home residence, and distance from hospital as explanatory variables. Patient-level analysis of demography, deprivation, comorbidity, and acuity markers was subsequently performed on referred and clinically assessed acute medical patients (n = 42,424) to examine how practice referral behaviour reflects clinical need for inpatient hospital care.
Results
Variance in GP referral rates for acute medical assessment was high (2.53-fold variation 1st vs. 4th quartile) and incompletely explained by increasing age and deprivation (adjusted R2 0.67, P < 0.001) such that significant variance remained after correction for confounders (2.15-fold). Patients from the highest referring quartile were significantly less likely to require hospital admission than those from the third, second, or lowest referring quartiles (adjusted odds ratio 1.28 [1.21–1.36, P < 0.001]; 1.30 [1.23–1.37, P < 0.001]; 1.53 [1.42–1.65, P < 0.001]).
Conclusions
High variation in GP practice referral rate for acute medical assessment is incompletely explained by practice population socioeconomic factors and negatively associates with need for urgent inpatient care. Identifying modifiable factors influencing referral rate may provide opportunities to facilitate community-based care and reduce congestion on acute unscheduled care pathways.
Variation in general practice (GP) referral rates to outpatient services is well described however variance in rates of referral to acute medical units is lacking.
Objective
To investigate variance in GP referral rate for acute medical assessment and subsequent need for hospital admission.
Methods
A retrospective cohort study of acute medical referrals from 88 GPs in Lothian, Scotland between 2017 and 2020 was performed using practice population size, age, deprivation, care home residence, and distance from hospital as explanatory variables. Patient-level analysis of demography, deprivation, comorbidity, and acuity markers was subsequently performed on referred and clinically assessed acute medical patients (n = 42,424) to examine how practice referral behaviour reflects clinical need for inpatient hospital care.
Results
Variance in GP referral rates for acute medical assessment was high (2.53-fold variation 1st vs. 4th quartile) and incompletely explained by increasing age and deprivation (adjusted R2 0.67, P < 0.001) such that significant variance remained after correction for confounders (2.15-fold). Patients from the highest referring quartile were significantly less likely to require hospital admission than those from the third, second, or lowest referring quartiles (adjusted odds ratio 1.28 [1.21–1.36, P < 0.001]; 1.30 [1.23–1.37, P < 0.001]; 1.53 [1.42–1.65, P < 0.001]).
Conclusions
High variation in GP practice referral rate for acute medical assessment is incompletely explained by practice population socioeconomic factors and negatively associates with need for urgent inpatient care. Identifying modifiable factors influencing referral rate may provide opportunities to facilitate community-based care and reduce congestion on acute unscheduled care pathways.
| Original language | English |
|---|---|
| Journal | Family Practice |
| Early online date | 5 Sept 2022 |
| DOIs | |
| Publication status | E-pub ahead of print - 5 Sept 2022 |