Background Deaths in HIV-positive people have decreased since the introduction of highly active antiretroviral therapy (HAART) in 1996. Fewer AIDS-related deaths and an ageing cohort have resulted in an increase in the proportion of HIV patients dying from non-AIDS-related disorders. Here we describe mortality and causes of death in people diagnosed with HIV in the HAART era compared with the general population. Methods In this observational analysis, we linked cohort data collected by Public Health England (PHE) for individuals aged 15 years and older, diagnosed with HIV in England and Wales from 1997 to 2012, to the Office for National Statistics (ONS) national mortality register. Cohort inclusion began at diagnosis with follow-up clinical information collected every year from all 220 National Health Service (NHS) HIV outpatient clinics nationwide. To classify causes of death we used a modified Coding Causes of Death in HIV (CoDe) protocol, which uses death certificate data and clinical markers. We applied Kaplan-Meier analysis for survival curves and mortality rate estimation and Cox regression to establish independent predictors of all-cause mortality, adjusting for sex, infection route, age at diagnosis, region of birth, year of diagnosis, late diagnosis, and history of HAART. We used standardised mortality ratios (SMRs) to make comparisons with the general population. Findings Between 1997 and 2012, 88 994 people were diagnosed with HIV, contributing 448 839 person-years of follow up. By the end of 2012, 5302 (6%) patients had died (all-cause mortality 118 per 10 000 person-years, 95% CI 115-121). In multivariable analysis, late diagnosis was a strong predictor of death (hazard ratio [HR] 3.50, 95% CI 3.13-3.92). People diagnosed more recently had a lower risk of death (2003-07: HR 0.66, 95% CI 0.62-0.70; 2008-12: HR 0.65, 95% CI 0.60-0.71). Cause of death was determinable for 4808 (91%) of 5302 patients; most deaths (2791 [58%] of 4808) were attributable to AIDS-defining illnesses. Cohort mortality was significantly higher than the general population for all causes (SMR 5.7, 95% CI 5.5-5.8), particularly non-AIDS infections (10.8, 9.8-12.0) and liver disease (3.7, 3.3-4.2). All-cause mortality was highest in the year after diagnosis (SMR 24.3, 95% CI 23.4-25.2). Interpretation Despite the availability of free treatment and care in the UK, AIDS continues to account for the majority of deaths in HIV-positive people, and mortality remains higher in HIV-positive people than in the general population. These findings highlight the importance of prompt diagnosis, care engagement, and optimum management of comorbidities in reducing mortality in people with HIV. Copyright (C) The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY license.
