Community based study of treatment seeking among subjects with symptoms of sexually transmitted disease in rural UgandaBMJ 1998; 317 doi: https://doi.org/10.1136/bmj.317.7173.1630 (Published 12 December 1998) Cite this as: BMJ 1998;317:1630
- L A Paxton (), assistant (clinical) directora,
- N Kiwanuka, field medical directorb,
- F Nalugoda, field supervisorb,
- R Gray, professorc,
- M J Wawer, associate (clinical) professora,
- Group Rakai Project Study
- aCenter for Population and Family Health, Columbia University School of Public Health, New York, NY 10032, USA
- bRakai Project, Uganda Virus Research Institute, Entebbe, Uganda
- cJohns Hopkins University School of Hygiene and Public Health, Baltimore, MD 21205, USA
- Correspondence to: Dr L A Paxton, Centers for Disease Control and Prevention, Mailstop E45, 1600 Clifton Road, Atlanta, GA 30333, USA
Two important impediments to the control of sexually transmitted diseases in Africa are unrecognised infections and ineffective treatment of infected people with symptoms.1 We report on treatment seeking among people with symptoms in a cohort in rural Uganda and quantify the extent to which asymptomatic infections and treatment seeking behaviour may affect control of sexually transmitted disease.
Subjects, methods, and results
The methods and preliminary results of the Rakai STD control for AIDS prevention study have been described in detail elsewhere.2 Adults enrolled in the initial survey provided specimens (blood, urine, vaginal swabs) to determine baseline prevalence of infection. They were allocated to intervention or control group according to community. At follow up 10 months later the behaviour of people experiencing symptoms between surveys was documented to estimate the proportion of symptomatic infected people expected to seek treatment. We then fitted both baseline prevalence of infection and treatment seeking between surveys to the Piot-Fransen STD control model.3
Serum samples were tested for syphilis (TRUST, New Horizons Diagnostics; TPHA, Sero-Tek, Miles Diagnostics) and urine for Neisseria gonorrhoeae and Chlamydia trachomatis by using ligase chain reaction (Abbott Lcx Probe System, Abbott Laboratories). Vaginal swabs were used for Trichomonas vaginalis culture (InPouch TV, BioMed Diagnostics) and Gram stain diagnosis of bacterial vaginosis. All respondents with syphilis in the intervention arm received penicillin treatment in the home while those in the control arm were counselled and referred to local health centres for free treatment. All subjects with possible symptoms of a sexually transmitted disease between study visits were advised to seek treatment at local health centres or project clinics.
In the first survey 12 827 people were enrolled and provided samples from which the baseline prevalence of infection was determined (syphilis 10.0%, gonorrhoea 1.6%, chlamydia 3.1%, and, in women, trichomonas 24.3%, bacterial vaginosis 49.9%). Most were asymptomatic: 53.3% of men and 65.6% of women with gonorrhoea, 91.7% of men and 76.0% of women with chlamydia, and 81.0% of women with trichomonas or bacterial vaginosis reported no symptoms in the previous 6 months.
Of the people initially enrolled, 9662 were seen at follow up. In the 10 months between the first and second surveys, 30.4% of women and 9.7% of men experienced genital tract symptoms. Rates were similar in both study arms. Women commonly reported vaginal itching (15.4%), pelvic pain (14.7%), vaginal discharge (9.4%), and genital ulcer (5.5%). Men reported genital ulcer (4.3%), dysuria (4.2%), and urethral discharge (1.7%). Over 40% reported they had done nothing to treat symptoms or prevent transmission and less than 40% of subjects in the control arm who were counselled had sought treatment for syphilis. Overall, 59.1% of symptomatic men and 55.4% of women sought treatment, and only 16.9% and 3.9%, respectively, notified their partners.
Almost three quarters of those seeking treatment used government health centres or private clinics; the rest treated themselves or visited traditional healers. More men than women chose to treat themselves (27.5% v 11.3%), while women more commonly resorted to traditional healers (14.7% v 5.1%). More than half reported sexual intercourse, and only 4.5% of men and 0.5% of women used condoms while they had symptoms.
Despite studies showing considerable deficiencies in care of people with sexually transmitted diseases within the formal health care sector,4 to fit the Piot-Fransen model we assumed that people seen at clinics or health centres receive adequate care and that those who treat themselves or use traditional remedies do not. We assumed that people without symptoms do not seek treatment at all and found that less than 8% of infected people in this cohort received adequate treatment.
In this study of treatment seeking for sexually transmitted diseases we have shown that relying on the treatment of only those with symptoms would reach only a small proportion of the infected population and that many would be unlikely to receive effective care. Thus, sexually transmitted disease control programmes in medically underserved populations must take into account both the prevalence of asymptomatic infections and the current health related practices of people with symptoms to design appropriate strategies to reduce transmission of these diseases.
Members of the study group were N Sewankambo, D Serwadda, F Wabwire Mangen, D McNairn, T Lutalo, F Makumbi, and M Meehan.
Contributors: LAP was the resident epidemiological adviser, oversaw the execution of the project, and helped with study design and analysis. NK was trial medical officer and helped in implementing the project, in field work, and in collecting data. FN was trial field supervisor and helped develop data collection and quality conrol procedures and oversaw them. RG was co-principal investigator and contributed to study design, implementation, and data analysis. MJW was the principal investigator, helped with study design, implementation, execution, and data analysis, and is guarantor for the study. N Sewankambo was Uganda principal investigator and was responsible for study design, implementation, and data interpretation. D Serwadda was Uganda co-principal investigator and was responsible for study design, implementation, monitoring, and data interpretation. D McNairn and M Meehan coordinated and supervised in-country laboratory activities. T Lutalo and F Makumbi were data managers and contributed to data analysis and interpretation.
Funding: This study was supported by grant RO1AI34826 from the National Institutes of Allergy and Infectious Diseases and by grant 5P30HD06826 from the National Institute of Child Health and Development, US National Institutes of Health. Additional support was given by the Rockefeller Foundation and the World Bank's Uganda STI project.
Competing interests: None declared.