Survival of Persons with and without HIV Infection in Denmark, 19952005

Context The expected survival of HIV-infected patients has been difficult to measure by comparing selected HIV samples and control groups. Contribution Denmark carefully tracks each of its residents' vital status, which makes it possible to accurately compare survival of HIV-infected persons and uninfected persons. In 2000 to 2005, life expectancy at age 25 years was 51 years in the general population and 39 years for HIV-infected persons without hepatitis C virus infection (HCV). Cautions Denmark provides excellent access to HIV and HCV care, so the results may be atypical. Implications Persons with HIV infection have a good, but far from normal, life expectancy. The Editors Knowing the expected survival of HIV-infected patients is of major public health interest. Mortality rates have decreased substantially in recent years as a result of improved effectiveness of highly active antiretroviral therapy (HAART) (1). Studies comparing mortality rates for HIV-infected persons with age- and sex-specific mortality rates for the general population (25) have reported 3- to 10-fold increase in successfully treated patients. The relative mortality rate, however, is highly dependent on the age distribution of the study sample and does not in itself answer questions about survival. We therefore aimed to estimate median survival and age-specific mortality rates for an entire HIV-infected population compared with a cohort from the general population. Persons with HIV infection were followed from before initiation of HAART and included those with such predictors of lower survival as poor response to therapy, AIDS diagnosis, low CD4 count, high viral load, and poor adherence to treatment (6, 7). Linking data from the population-based Danish HIV Cohort Study (DHCS) (8) and the Danish Civil Registration System (CRS) (9, 10) allowed us to use product-limit methods that are analogous to the period life tables used by national authorities for estimating median survival (11). Methods Study Sample The DHCS is a prospective, nationwide, population-based cohort study of all HIV-infected persons treated in Danish HIV clinics since 1 January 1995 (8, 12). The study is ongoing, with continuous enrollment of both newly diagnosed residents and immigrants with existing HIV infection. Treatment for HIV infection in Denmark is restricted to 8 specialized centers, and the Danish health care system provides free tax-supported medical care, including antiretroviral treatment for HIV infection. The study databases are updated annually. Adult (>16 years) DHCS participants with residency in Denmark were included at their first visit to an HIV clinic. The Civil Registration System (CRS) is a national registry of all Danish residents; this registry contains information on date of birth; sex; address; date of migration; and date of death, if applicable (9). A 10-digit personal number (Central Person Registry [CPR] number), assigned at birth, uniquely identifies each person. The CRS is updated within a week of a person's birth, address change, death, or emigration. Use of the CPR number enables treatment centers to avoid multiple registrations of the same patient and allows tracking of deaths and persons lost to follow-up due to emigration. Using the CRS records, we drew a random sample of persons from the general population and matched them to each HIV-infected patient according to sex and month of birth and residence in the same municipality as the patient on the date of diagnosis (Denmark has a population of approximately 5.3 million persons living in 270 municipalities). We aimed to sample 99 persons from the general population for each HIV-infected person. However, because of a shortage of eligible persons in some municipalities, the mean number of persons from the general population per patient was 95.2 in the final sample. Patients with at least 1 positive result on a hepatitis C virus (HCV)antibody test or a positive result on an HCV RNA test were considered to be HCV-positive; the other patients were considered to be HCV-negative. The HCV-antibody status was available for 88.4% of all patients and for 95.4% of patients observed during 2000 to 2005. We did not have individual data on HCV infection in the general population, but the estimated prevalence in Denmark is only 3 per 1000 (13). Highly active antiretroviral therapy was defined as the combination of antiretroviral treatment with at least 3 drugs, including at least 1 protease inhibitor, 1 nonnucleoside reverse transcriptase inhibitor, or abacavir; or the 2-drug combination of efavirenz- and ritonavir-boosted lopinavir. Treatment interruption was defined as a period of at least 2 weeks after initiation of HAART during which the patient did not take antiretroviral drugs. Structured treatment interruptions have generally not been recommended in Denmark. Causes of death extracted from patient files were available for patients in DHCS and were divided into HIV-related causes (AIDS-defining illnesses and bacterial infections, corresponding to International Classification of Disease, Tenth Revision [ICD-10], codes A02, A07.2-07.3, A15-19, A31, A81.2, B00, B20-25, B37-39, B45, B58, C46, C53, C83.4, C83.9, F02.4, and J13-17), nonHIV-related other causes, and unknown causes. Statistical Analysis We computed KaplanMeier life tables using age as the time scale. Persons with HIV infection were observed from the date of HIV diagnosis or from the first visit to an HIV clinic, if the visit occurred at a later date; persons from the general population were observed starting at the same time as their matched HIV-infected patients. All persons were censored at emigration or on 1 May 2005. Death from any cause was the outcome event. We estimated median survival times and computed mortality rates from age 25 years separately for men and women and for the subgroup of HCV-negative persons. We chose 25 years because this was the youngest age group with a sufficient number of patients (n= 170) being observed. We performed analyses for 3 clinically relevant periods: 1995 to 1996 (pre-HAART), 1997 to 1999 (early HAART), and 2000 to 2005 (late HAART) and with respect to the length of HAART treatment: before HAART, first year, second and third years, fourth and fifth years, and sixth year onward. We computed mortality rates in 5-year age intervals and estimated crude relative (mortality rate ratio [mortality rate for patients with HIV infection divided by the mortality rate for the general population]) and absolute (excess mortality rate [mortality rate for the general population subtracted from the mortality rate for patients with HIV infection]) effects in HIV-infected patients compared with persons from the general population. In accordance with the matched design, we used a stratified Cox regression model for the mortality rate ratio estimations. When comparing mortality rate ratios from the stratified model with crude mortality rate ratios, we found only small deviations (median, 4.3% [interquartile range, 1.2% to 7.0%]) and therefore used standard statistical methods for the excess mortality rates. We further examined mortality rate ratios and excess mortality rates for the late HAART period and for HCV-negative and HCV-positive persons during the late HAART period. Data analysis was performed by using Stata statistical software, version 9.0 (Stata Corp., College Station, Texas). Approvals and Permissions The Danish Data Protection Agency approved the establishment of the cohort study. The study was not subject to approval by the ethics committee because data collection did not involve direct patient contact. Role of the Funding Sources The Danish HIV Cohort study receives funding from the Danish AIDS Foundation, Odense University Hospital, Preben and Anna Simonsen's Foundation, the Foundation of the Danish Association of Pharmacists, and the Clinical Institute at the University of Southern Denmark. The funding sources were not involved in the design, data collection, analysis, or writing of the study. Results Study Sample We included 3990 HIV-infected persons and 379872 persons from the general population: The respective median observation time after age 25 years was 5.8 person-years (interquartile range, 2.2 to 9.9 person-years) and 8.4 person-years (interquartile range, 4.3 to 10.3 person-years), respectively (Table 1). One hundred twenty-one (3.0%) HIV-infected patients and 11552 (3.0%) persons from the general population were lost to follow-up; of these, 107 (2.7%) patients with HIV infection and 10234 (2.7%) persons from the general population emigrated. There were 2045 (51%) incident HIV cases diagnosed after 1 January 1995; 75% were observed within 31 days of diagnosis, and 95% came under observation within the first 181 days after diagnosis. After HAART was introduced in 1996, the prevalence of patients receiving this treatment gradually increased, surpassing 75% in 2002 to 2004. At any given time, fewer than 5% of HIV-infected patients were interrupting treatment. The number of patients under observation varied with age and was highest (range, 515 to 1004) for those who were 30 to 50 years of age (Table 2). Table 1. Characteristics of Study Participants* Table 2. Age-Specific Mortality Rates* Survival from Age 25 Years All participants were observed from age 25 years: HIV-infected persons had a median survival of 19.9 years (17.5 years for men and 24.2 years for women), whereas persons from the general population had a median survival of 51.1 years (50.8 years for men and 54.8 years for women) (Table 3). During the late HAART period (2000 to 2005), median survival of HIV-infected patients had increased to 32.5 years (32.1 years for men and 32.3 years for women) overall, and to 38.9 years (37.8 years for men and 40.1 years for women) after persons with known HCV infection were excluded (Figure). Table 3. Median Survival and Mortality Rates Starting at Age 25 Years* Figure. Survival from age 25 years