Factors related to attrition in a cohort study of HIV in Malawi.

BACKGROUND: Longitudinal studies face power reduction due to loss to follow up (LTFU). Bias may also arise because of differences between those who stay in the study and those who are LTFU. We studied factors associated with LTFU in a cohort of HIV seronegative and sera-positive mothers in urban Malawi. OBJECTIVE: To bridge the existing gaps by examining the factors associated with attrition. DESIGN: Longitudinal study. SETTING: Queen Elizabeth Central Hospital (QECH) and the Kamuzu Central Hospital in Blantyre, Malawi. SUBJECTS: One thousand three hundred and fifty three women who attended the prenatal clinic, between October 1989 and October 1990 were recruited as part of a study to determine rates and risk factors of sero-prevalence and sera-conversion of HIV-1 among this cohort. RESULTS: In this cohort study, 1353 women were enrolled at delivery and 1188 (88%) returned for the first follow-up visit at three months post-partum. Of those who returned, 177 (15%) were subsequently lost during the remaining months of follow-up. The main predictors of LTFU were younger maternal age, lower educational level of the father, HIV infection of the mother, lower birth weight of the index child and mother not being married. CONCLUSIONS: Researchers planning studies in developing countries should consider the impact of lower education and poorer infant health on study retention in developing countries.

Transmission of Mycobacterium tuberculosis through casual contact with an infectious case.

BACKGROUND: An ongoing restriction fragment length polymorphism study of Mycobacterium tuberculosis isolates from tuberculosis cases showed an identical 12-band IS6110 pattern unique to 3 unrelated patients (Patients A-C) diagnosed as having tuberculosis within a 9-month period. METHODS: In an attempt to identify epidemiologic links between the 3 patients, we performed site visits to the retail business work site of patient A and conducted detailed interviews with all 3 patients and their contacts. RESULTS: Patient B had visited patient A's work site 3 times during patient A's infectious period, spending no more than 15 minutes each time. Patient C visited patient A's work site on 6 to 10 occasions during this period for no more than 45 minutes at any one time. There were no other epidemiologic links between these 3 cases other than the contact at the store. Contact investigation identified 4 tuberculin skin test conversions among 8 (50%) of patient A's coworkers, 6 positive tests among 15 household contacts (40%), and 8 positive tests among 16 identified customers who were casual contacts (50%). Patient B and patient C were most likely infected by patient A during one of their brief visits to patient A's work site. CONCLUSIONS: These data demonstrate that some tuberculosis is spread through casual contact not normally pursued in traditional contact investigations and that, in certain situations, M tuberculosis can be transmitted despite minimal duration of exposure. In addition, this outbreak emphasizes the importance of DNA fingerprinting data for identifying unusual transmission in unexpected settings.

Ventriculitis complicating use of intraventricular catheters in adult neurosurgical patients.

Ventriculitis is a serious complication of intraventricular catheter (IVC) use, with rates of IVC-related infections ranging from 0% to 45% and gram-positive organisms predominating. We prospectively analyzed ventriculostomy-related infections occurring among 157 adult neurosurgical patients (mean age, 54.9 years; 90 [57%] were women) from 1995 through 1998, to determine the incidence of, risk factors for, and organisms that cause ventriculitis. A total of 196 IVC events resulted in 11 infections (5.6%; 9 were caused by gram-negative organisms and 2 by coagulase-negative staphylococci). Independent risk factors for IVC-related infection include length of IVC placement (8.5 days [infected] vs. 5.1 days [uninfected]; P=.007) and cerebrospinal fluid leakage about the IVC (P=.003). The length of hospital stay (30.8 days vs. 22.6 days; P=.03) and mean total hospital charges ($85,674.27 vs. $55,339.21; P=.009) were greater for infected patients than for uninfected patients. In addition, a microbiologic shift from gram-positive organisms toward gram-negative organisms was noted. This study suggests that IVC-related infections remain serious infections that increase the length of hospitalization.

An outbreak of scabies in a teaching hospital: lessons learned.

OBJECTIVE: To investigate an outbreak of scabies in an inner-city teaching hospital, identify pathways of transmission, institute effective control measures to end the outbreak, and prevent future occurrences. DESIGN: Outbreak investigation, case-control study, and chart review. SETTING: Large tertiary acute-care hospital. RESULTS: A patient with unrecognized Norwegian (crusted) scabies was admitted to the acquired immunodeficiency syndrome (AIDS) service of a 940-bed acute-care hospital. Over 4 months, 773 healthcare workers (HCWs) and 204 patients were exposed to scabies. Of the exposed HCWs, 147 (19%) worked on the AIDS service. Risk factors for being infested with scabies among HCWs included working on the AIDS service (odds ratio [OR], 5.3; 95% confidence interval [CI95], 2.17-13.15) and being a nurse, physical therapist, or HCW with extensive physical contact with infected patients (OR, 4.5; CI95, 1.26-17.45). Aggressive infection control precautions beyond Centers for Disease Control and Prevention barrier and isolation recommendations were instituted, including the following: (1) early identification of infected patients; (2) prophylactic treatment with topical applications for all exposed HCWs; (3) use of two treatments 1 week apart for all cases of Norwegian scabies; (4) maintaining isolation for 8 days and barrier precautions for 24 hours after completing second treatment for a diagnosis of Norwegian scabies; and (5) oral ivermectin for treatment of patients who failed conventional therapy. CONCLUSIONS: HCWs with the most patient contact are at highest risk of acquiring scabies. Because HCWs who used traditionally accepted barriers while caring for patients with Norwegian scabies continued to develop scabies, we found additional measures were required in the acute-care hospital. HCWs with skin exposure to patients with scabies should receive prophylactic treatment. We recommend (1) using heightened barrier precautions for care of patients with scabies and (2) extending the isolation period for 8 days or 24 hours after the second treatment with a scabicide for those patients with Norwegian scabies. Oral ivermectin was well tolerated for treating patients and HCWs who failed conventional treatment. Finally, we developed a surveillance system that provides a "barometric measure" of the infection rate in the community. If scabies increases in the community, a tiered triage system is activated to protect against transmission among HCWs or hospital patients.

Errors in the treatment of tuberculosis in Baltimore.

BACKGROUND: Incomplete or incorrect antibiotic therapy, especially in the initial phase of antituberculosis (anti-TB) treatment, is a major cause of acquired drug resistance and treatment failure. We determined the extent of errors in anti-TB treatment regimens by way of nonadherence to recommended treatment protocols among patients with TB in Baltimore, MD, a city with declining rates of disease. An error was defined as using too few drugs or the wrong drugs, giving inadequate doses of drugs, or prescribing an inadequate duration of treatment. METHODS: We reviewed the records of all patients with culture-positive, pulmonary TB reported in the city of Baltimore from January 1, 1994, to December 31, 1995. We determined demographic information, initial anti-TB regimen, doses and duration of therapy, history or presence of resistance to anti-TB drugs, injecting-drug or alcohol abuse, HIV status, and whether treatment was given by a private physician or by the Tuberculosis Clinic of the Baltimore City Health Department (BCHD). RESULTS: Of the 110 cases of active pulmonary TB, 17 cases (15.4%) had errors in treatment for control of their current disease. Thirteen of 34 privately treated patients (38%) had some error in their initial anti-TB regimen, compared with 4 of 76 patients (5.2%) treated by the Tuberculosis Clinic of the BCHD (p < 0.0001). Patients were otherwise similar as determined by age, sex, HIV status, drug-resistance characteristics, and injecting-drug use, regardless of whether they had erroneous anti-TB regimens. CONCLUSION: In a low-prevalence area, private physicians make frequent errors in prescribing anti-TB therapy. Additional educational resources for physicians and increased use of expert consultation may contribute to improved TB control.