Loss to follow-up in HIV immigrant patients. Is it a relevant problem? Results of the Spanish COMESEM cohort

Purpose of the study

To assess epidemiological characteristics and loss to follow-up in immigrant patients, and compare them with Spanish-born patients, in two periods: 1983–1999 and 2000–2008.

Methods

COMESEM cohort is a prospective, population-based HIV-infected patient cohort of the south and east Greater Madrid (Spain). To analyze "loss to follow-up" of our patients, we collected the following data from our databases: age, year of diagnosis, gender, date of the last visit, risk group and AIDS diagnosis. We considered as "loss to follow-up", patients who were alive in their last visit, had not requested transfer to another medical facility and were not registered as dead in the Mortality Registry of the Autonomous Community of Madrid. It was also required that at least 6 months had elapsed since their last visit. Loss to follow-up of immigrant and Spanish-born patients were compared by means of Kaplan-Meyer curves (long-rank test).

Summary of results

Of 8,166 patients in our database, 5,450 were amenable to analysis (852 were immigrants). Since there were statistically significant differences in the percentage of immigrant patients between both time periods (3.8% vs. 43.5%, p = 0.0001), we carried out the analysis by stratifying into the periods. There were no significant differences in age between time periods. In immigrant patients: the percentage of women increased in the second period (40% vs. 52%, p = 0.0003), so did the sexual transmission risk group (67.6% vs. 90.3%), and AIDS diagnoses in this population diminished (37.4% vs. 24.7%, p = 0,0001). Although there was statistically significant more "loss to follow-up" of immigrants than Spanish-born patients in the second period (p = 0.0004), this difference does not seem clinically important (11.9% in immigrant vs. 7.12% in Spanish-born patients).

Conclusion

1. In the last 8 years, around 50% of new diagnoses of HIV infection have been established in immigrant patients. 2. The frequency in immigrant patients has increased in women and in the sexually transmitted risk group. 3. AIDS diagnoses in this population have diminished in later years. 4. Although we observed differences in "loss to follow-up" between immigrants and Spanish-born patients, these do not seem very relevant. A likely explanation is the universal access to medical care in Spain, and particularly HIV treatment, which is also assured for immigrant patients.