The study was conducted from January 2007 to December 2008 as part of an effort to control and prevent viral hepatitis in the province. Four vulnerable or at risk population groups were identified for this cross-sectional study. These included IV drug users, prisoners, security personnel and healthcare workers. Health care workers were included in high risk groups due to higher exposure to hepatitis patients, routinely reported needle prick injuries and low uptake of hepatitis vaccine. This group is of particular importance because, infected health care workers can transmit the infection to uninfected patients admitted to hospital, which can further spread the disease in society.
We compiled a database of prisoners in Landhi Jail, health care workers of the Civil Hospital Karachi, IDUs registered by the Referral Laboratory of Sindh AIDS Control Program and security personnel from a large private company. A unique identification number was assigned to all individuals in this database. A computerized program was run to generate a random sample of the individuals included in this database. All identified individuals were then approached for a detailed interview and blood sample collection. The response rate was 83% in this study with a total of 4202 individuals included in this analysis.
We estimated the sample size to measure a 2% difference of hepatitis prevalence (assuming 4% prevalence) [5, 6] between groups at 0.05 significance level using a two-sided comparison and the power of 90%. Sample size was computed for both the chi-squared test using the Yates’ continuity correction and Fisher Exact test. A sample of 3252 participants was the minimum number required to be accrued in order to perform this survey.
Among HCWs, only staff from exposure-prone procedures (EPP) was included in the study. These included intensive care unit staff, surgeons, nursing staff, phlebotomists, doctors, nurses and other paramedical staff providing direct care to the patient. Demographic information of screened people was recorded along with history of risk factors. A team of laboratory workers was deputed to collect samples from these groups. All samples were taken after obtaining the informed consent. Consent was also obtained from the institutional heads with assurance of confidentiality and provision of treatment from the hepatitis control program. The study protocol was reviewed and approved by an independent ethics committee. Sample collection was carried out at different sites for the groups as indicated. Five milliliter of venous blood was collected in gel vaccutainer tubes (yellow top) using aseptic technique. Blood samples were transported to the designated lab for further processing.
HBsAg and HCV Antibody tests were performed at the central lab of the Civil Hospital Karachi (CHK) using the CMEIA (Chemoluscent micro particle enzyme immunoassay) method on blood samples from HCW, prisoners and security personnel. Samples from IDUs were tested at the Referral laboratory of Sindh AIDS Control Program by ELISA Methods using Bio-Rad Monolisa Antigen-antibody Coombo test kits.
Data were recorded and analyzed using the Stata Software Version 11 (StataCorp, College Station, TX, USA). We categorized the age using 10-year groups from 21–30, 31–40, and 41–50 years old. The means between groups were compared using the independent sample t test and analysis of variance (ANOVA) with bonferroni adjustments for multiple comparisons was done. We explored the association between age group, gender and risk group by using the logistic regression model. Health-care worker group was taken as a reference category when investigating the relationship between risk group and hepatitis B and C. As this group may have prevalence of hepatitis B and C approximate to the background population or may be slightly higher than that. We constructed two different models for hepatitis B and Hepatitis C. Age was found to be significant on univariate analysis and was therefore included in the final multivariate model to make appropriate adjustments. The level of significance was set at 95% (p value < 0.05).