BackgroundDrug overdose is a leading cause of mortality and morbidity amongst people who inject drugs (PWID). Drug overdose surveillance typically relies on the International Classification of Diseases (ICD-10) coding system, however its real world utilisation and the implications for surveillance have not been well characterised. This study examines the patterns of ICD-10 coding pertaining to drug overdoses within emergency departments for a cohort of known PWID.MethodsCohort data from 688 PWID was linked to statewide emergency department administrative data between January 2008 and June 2013. ICD-10 diagnostic codes pertaining to poisonings by drugs, medicaments and biological substances (T-codes T36-T50) as well as mental and behavioural disorders due to psychoactive substance use (F-codes F10-F19) were examined.ResultsThere were 449 unique ED presentations with T or F code mentions contributed by 168 individuals. Nearly half of the T and F codes used were non-specific and did not identify either a drug class (n = 160, 36%) or clinical reaction (n = 46, 10%) and 8% represented withdrawal states. T and F codes could therefore be used to reasonably infer an illicit drug overdose in only 42% (n = 188) of cases. Majority of presentations with T or F overdose codes recorded only one diagnostic code per encounter (83%) and representing multiple-drug overdose (F19.- = 18%) or unidentified substances (T50.9 = 17%) using a single, broad diagnostic code was common.ConclusionsReliance on diagnoses alone when examining ED data will likely significantly underestimate incidence of specific drug overdose due to frequent use of non-specific ICD-10 codes and the use of single diagnostic codes to represent polysubstance overdose. Measures to improve coding specificity should be considered and further work is needed to determine the best way to use ED data in overdose surveillance.

Background People who inject drugs (PWID) have been identified as frequent users of emergency department (ED) and hospital inpatient services. The specific challenges of record linkage in cohorts with numerous administrative health records occurring in close proximity are not well understood. Here, we present a method for patient-specific record linkage of ED and hospital admission data for a cohort of PWID. Methods Data from 688 PWID were linked to two state-wide administrative health databases identifying all ED visits and hospital admissions for the cohort between January 2008 and June 2013. We linked patient-specific ED and hospital admissions data, using administrative date-time timestamps and pre-specified linkage criteria, to identify hospital admissions stemming from ED presentations for a given individual. The ability of standalone databases to identify linked ED visits or hospital admissions was examined. Results There were 3459 ED visits and 1877 hospital admissions identified during the study period. Thirty-four percent of ED visits were linked to hospital admissions. Most links had hospital admission timestamps in-between or identical to their ED visit timestamps (n = 1035, 87%). Allowing 24-h between ED visits and hospital admissions captured more linked records, but increased manual inspection requirements. In linked records (n = 1190), the ED ‘departure status’ variable correctly reflected subsequent hospital admission in only 68% of cases. The hospital ‘admission type’ variable was non-specific in identifying if a preceding ED visit had occurred. Conclusions Linking ED visits with subsequent hospital admissions in PWID requires access to date and time variables for accurate temporal sorting, especially for same-day presentations. Selecting time-windows to capture linked records requires discretion. Researchers risk under-ascertainment of hospital admissions if using ED data alone.

Introduction Risky drinking frequently remains undiagnosed or untreated, including in hospitalised inpatients. Using the Alcohol Use Disorders Identification Test (AUDIT), we assessed the feasibility of screening for risky drinking and whether screening results aligned with alcohol‐attributable diagnoses in an inpatient population. Methods We conducted a cross‐sectional survey across a tertiary health service in Melbourne, Australia. Researchers collected demographics, AUDIT scores and acceptability from all eligible adult inpatients available on day of survey. Main outcomes were prevalence of risky drinking (AUDIT ≥8), mean AUDIT score and patient acceptability. Identification of risky drinking by the abbreviated ‘AUDIT‐C’ or discharge diagnoses (extracted by data‐linkage with medical records) was compared. Results Of 473 eligible inpatients, 61% (n = 289) participated, 22% (n = 103) were unavailable and 17% (n = 81) declined. Median age was 64 years (IQR = 48, 76); 54% (n = 157) were male. Mean AUDIT score was 4.4 (SD = 5.5). Risky drinking prevalence was 20% (n = 57), 2% (n = 7) had scores suggestive of dependence (AUDIT ≥20, a subset of risky drinkers). Odds of risky drinking were reduced in females (OR 0.19, 95% CI 0.09, 0.41; P < 0.001) and participants ≥70 years (OR 0.22, 95% CI 0.07, 0.71; P = 0.01). Alcohol‐attributable diagnoses did not consistently align with risky drinking, with half of inpatients with wholly attributable diagnoses classified as low risk. Most inpatients considered screening acceptable (89%, n = 256). Discussion and Conclusions Pre‐admission risky drinking was evident in one‐fifth of hospital inpatients, but alcohol‐attributable diagnoses were unreliable proxy measures of risky drinking. Screening in‐patients with the AUDIT was acceptable to inpatients and can be feasibly implemented in an Australian tertiary hospital setting.