Article Text
Statistics from Altmetric.com
The article by Grassly et al shows the impact of errors in the national data on HIV/AIDS estimates.1 We would like to present our experience with the errors in the official records in Iran that may confound the estimation and expert judgment.
The results of our of HIV/AIDS preventive education in the schools in Iran were released in December 2003.2 To do the study, we had to review the only reliable official report in the country—namely, the registry of the Centre for Disease Control (CDC), Ministry of Health and Medical Education (MOHME) of Iran.3 Figure 1 depicts the trend of the new cases of HIV/AIDS found in Iran in recent years according to the CDC registry. While reviewing the report with our colleagues in the UNESCO HIV/AIDS Coordination Unit, we noticed fluctuations in the data for the years 1996 and 2001; there are peaks in 1996 and 2001. We checked the data obtained from the CDC and found the data to be in accordance with their records. We then asked about the setting of data collection and found the reason for the peaks. In 1996 and 2001, the prisons throughout the country were tested for HIV by the MOHME and the data were included in the annual reports of the CDC.
Trend of the new cases of HIV and AIDS patients according to the registry of the CDC, MOHME of Iran (1986–2003). Figure 1 The data for the year 2003 covers the number reported cases of HIV/AIDS until September.
Both trafficking and using of drugs are crimes in Iran, and every year large numbers of this high risk population are imprisoned on drug related charges. Records from the CDC show that injecting drug use is the main source of transmission of HIV in Iran (62.78 %), and sexual contact is said to be the second most important route of transmission, 7.27 %3; 26.12% of the cases, however, are grouped as “unspecified route of transmission.”3 This is as long as there is no official record on prevalence of prostitution in Iran.4 The MOHME has recently announced an estimated figure of 137 000 injecting drug users in Iran.5 We should be careful as the statistics from places of concentrated epidemics (for example, prisons) can bias the results and more data are needed to process an accurate estimation of the real number of infected people.
The growing number of new cases can also be because of both the advance in diagnostic methods and the increasing number of the infected people. New cases reported by the Blood Transfusion Organisation of Iran for example make a considerable number of new cases reported by CDC, most of whom are grouped as “unspecified route of transmission.”
Accurate data are a must in decision making while expert judgment is needed to interpret the results of the formal statistical inference.1 The data presented by the Iranian CDC does not seem to represent the whole population and so cannot be used for estimation of the real number of patients. Therefore, it is recommended that the CDC explain in detail the setting of the data collection in its periodic reports to safeguard against possible overestimation or underestimation of HIV/AIDS. This can help experts, and especially international agencies such as UNAIDS, to make more accurate estimates of HIV/AIDS. Expert judgment is needed in policy making for such important health issues and official statistics cannot always be trusted.