SETTING:Antiretroviral therapy (ART) reduces pulmonary tuberculosis (PTB) in human immunodeficiency virus (HIV) infected children. Recent ART recommendations have increased the number of children on ART.OBJECTIVE:To determine the prevalence and incidence of TB in HIV-infected children after the implementation of expanded ART guidelines.DESIGN:A prospective cohort study including HIV-infected children aged 6 weeks to 14 years was conducted in Kenya. The primary outcome measure was clinically diagnosed TB. Study participants were screened for prevalent TB at enrollment using Kenya's national guidelines and followed at monthly intervals to detect incident TB. Predictors of TB were assessed using logistic regression and Cox proportional hazards regression.RESULTS:Of 689 participants (median age 6.4 years), 509 (73.9%) were on ART at baseline. There were 51 cases of prevalent TB (7.4%) and 10 incident cases, with over 720.3 child-years of observation (incidence 1.4 per 100 child-years). Months on ART (adjusted hazard ratio [aHR] 0.91, P = 0.003; aOR 0.91, P< 0.001) and months in care before ART (aHR 0.87, P= 0.001; aOR 0.92, P < 0.001) were protective against incident and prevalent TB.CONCLUSIONS:ART was protective against TB in this cohort of HIV-infected children with high levels of ART use. Optimal TB prevention strategies should emphasize early ART in children.

Background: Antibiotic treatment is not recommended for acute bronchitis in immunocompetent patients in industrialised countries. Whether these recommendations are relevant to the developing world and to immunocompromised patients is unknown. Design, setting and participants: Randomised, triple blind, placebo controlled equivalence trial of amoxicillin compared with placebo in 660 adults presenting to two outpatient clinics in Nairobi, Kenya, with acute bronchitis but without evidence of chronic lung disease. Main outcome measure: The primary study end point was clinical cure, as defined by a ⩾75% reduction in a validated Acute Bronchitis Severity Score by 14 days; analysis was by intention to treat with equivalence defined as ⩽8% difference between study arms. Results: Clinical cure rates in the amoxicillin and placebo arms were 81.7% and 84.0%, respectively (difference 2.3%, 95% CI −8.6% to 4.0%). Of 131 HIV infected subjects (19.8%), cure rates for those randomised to amoxicillin (77.2%) and placebo (83.8%) differed by 6.6% (95% CI −21.7% to 8.6%). Among HIV uninfected subjects, the difference in cure rates was 1.6% (95% CI −8.5% to 5.3%). Potential drug side effects were similar in the two arms. No subjects required hospitalisation or died. Conclusion: Antibiotic treatment of acute bronchitis is unhelpful, even in populations with a high prevalence of HIV infection.

INTRODUCTION:Although several clinical prediction rules exist for lower respiratory tract infection (LRTI), few are for acute bronchitis (acute bronchitis) and most have not been validated in high human immunodeficiency virus (HIV) prevalence settings. METHODS:An Acute Bronchitis Severity Score (ABSS) was developed and validated during a randomized trial of antibiotic treatment for acute bronchitis. Ambulatory adults with productive cough of < or =2 weeks at out-patient respiratory disease clinics in Nairobi, Kenya, were recruited and assessed for clinical response to therapy. The ABSS quantitative ratings of LRTI-associated symptoms, physical signs and sputum Gram stain purulence were assessed using standard psychometric tests. RESULTS:The ABSS was evaluated among 649 cases of acute bronchitis; 129 (20%) were HIV-seropositive. The ABSS had small floor and ceiling effects (1.8/0.2) and demonstrated high internal consistency (alpha-coefficient of 0.66) and internal validity, with a mean inter item total correlation of > or =0.25. Effect sizes from baseline to subsequent follow-up visits were large (>0.5). Wheezing and chest pain were associated with higher ABSS values, whereas irrelevant clinical variables were not. CONCLUSION:The ABSS demonstrated good responsiveness, high internal consistency, good correlation with common respiratory signs and symptoms and high discriminatory validity among patients with acute bronchitis in a high HIV-seroprevalence setting.

SETTING:Risk factors for mortality in hospitalized patients with community-acquired pneumonia (CAP) are well known. There are limited data on prognostic indicators among out-patients.OBJECTIVE:To compare the clinical presentation, outcome and prognostic factors for clinical improvement in human immunodeficiency virus (HIV) infected and non-HIV-infected out-patients with CAP.METHODOLOGY:Adults in Nairobi with CAP were treated with erythromycin as first-line therapy. Clinical symptoms were evaluated using a validated CAP-related symptom score (CSS). Clinical improvement was defined as reduction of baseline CSS by > or = 50%.RESULTS:Of 531 adults enrolled with CAP, 422 (79.5%) completed follow-up. Participants had a mean age (+/- SD) of 33.7 +/- 11.4 years, 274 (51.6%) were male and 193 (37%) were HIV-seropositive with a higher baseline CSS (27 vs. 25, P < 0.006). Overall, 196 of 422 (46%) had clinical improvement by 28 days. Factors independently associated with a longer time to clinical improvement included not being married (adjusted hazard ratio [aHR] 0.66, 95% CI 0.48-0.92) and higher baseline CSS (aHR 1.05, 95% CI 1.03-1.06).CONCLUSIONS:HIV-infected and non-infected patients with CAP responded similarly to out-patient treatment, but HIV-infected patients were more likely to present with severe symptoms. Baseline CSS and marital status were predictive of time to clinical improvement.

Objective To assess the effect of steroid hormone contraceptive methods on the clinical course of early HIV infection among women in Kenya, Thailand and Zimbabwe. Methods Women with documented HIV infection and CD4 cell counts over 500 cells/μl who were asymptomatic or had early (mild disease) were invited to participate in an observational cohort study with 6-monthly follow-up visits for 4 years. Baseline information was collected through interview, physical examination and blood sample collection for CD4 cell count, HIV quantification and subtyping. Identical procedures were repeated at each follow-up visit. Study endpoints include HIV disease progression, the incidence of opportunistic infections (OI), and surrogate markers of disease progression, including changes in CD4 cell counts and viral load. These will be analysed according to the contraceptive methods used. The study aimed to recruit a total of 220 users each of DMPA, Norplant, COC and a similar number of women not using hormonal contraception over the study sites. Results Recruitment to the study started in 2000 and by the middle of 2002 a total of 395 women had been recruited, the majority of whom were from Nairobi (52%) and Harare (23%). One quarter of women were using COC, 42% DMPA, 5% Norplant and the remaining 28% non-hormonal contraception at enrolment. Almost all the Norplant users were from Bangkok and almost all the COC and DMPA users were from the two African sites. Few HIV endpoints have occurred in the cohort to date. CD4 cell counts at enrolment were lower in Harare than in Nairobi, despite the selection of women with counts above 500 cells/μl. Preliminary assessment of the rates of CD4 cell decline showed a higher rate in Harare (mean annual loss of 102 cells/μl; SD 244) than in Nairobi (mean annual loss of 55 cells/μl; SD 283). Recruitment into the cohort continued through 2003 and follow-up will continue for the enrolled women up to the scheduled four completed years. The study is estimated to have 80% power to demonstrate twofold differences in the rates of CD4 cell decline between oral contraceptive and non-hormonal users, as well as between DMPA and non-hormonal users. Conclusion Recruitment into the study has been difficult, particularly as the study was not able to offer any promise of therapy for HIV-positive women. Plans to provide antiretroviral therapy to those women whose CD4 cell counts decline below 200 cells/μl are under development and will allow the clinical response to antiretroviral therapy to be assessed in the well-documented cohort of steroid hormone contraception users. Recruitment to the study had to be abandoned in Brazil as the national standards for therapy meant that there would at best be only a short period of observation before the volunteers qualified for treatment, and it proved very difficult to identify HIV-positive women with CD4 cell counts over the study threshold for enrolment. Support Financial support for the study was provided by the World Health Organization Special Programme of Research, Development and Research Training in Human Reproduction (HRP), Geneva, Switzerland.

The purposes of this study were to measure incidence and determine risk factors associated with opportunistic infections (OIs) and mortality among an HIV-infected cohort in Nairobi, Kenya. Three hundred and eighty-one seropositive ambulatory adults in Nairobi, Kenya were followed from 1997 to 2000 with participants visiting the clinic every two months and when acutely ill. Acute bronchitis was the most frequent diagnosis, followed by sexually transmitted infections, candida vaginitis (among women), fever, diarrhoea, pneumonia, HIV-associated skin rash, oral candidiasis and urinary tract infection. Associations between the frequency of these diagnoses including survival and sociodemographic factors and initial CD4 count were assessed. A CD4 count <200 cells/mL at recruitment was strongly associated with decreased survival (adjusted odds ratio=3.0, 95% confidence interval 1.7-5.1). These findings may help to target high-risk populations and guide OI prevention and treatment strategies including decisions regarding initiation of antiretroviral therapy in sub-Saharan Africa.

Among women attending family planning clinics in Nairobi, Kenya, the HIV-seroprevalence rates for different contraceptive methods were: depomedroxyprogesterone acetate (DMPA) 431/3279 (13.1%), combination oral contraceptive pill 114/1073 (10.6%), and progesterone-only contraceptive pill (POCP) 45/741 (6.1%). After adjusting for age, marital status, and parity, women using the POCP had a lower HIV seroprevalence (adjusted odds ratio 0.5, 95% confidence interval 0.3-0.7) than women using DMPA. This association was most pronounced among POCP users of lower parity.

We evaluated a modified WHO algorithm for the management of chronic diarrhoea in ambulatory HIV-1-infected adults that included the use of norfloxacin, metronidazole and loperamide. In a cohort of 380 HIV-1-infected adults in Nairobi, Kenya, the incidence of chronic diarrhoea was 286 cases/1000 person-years; 85% of chronic diarrhoea episodes responded completely to therapy. Although these findings appear to validate the modified WHO chronic diarrhoea algorithm, randomized controlled trials may better define the indication for antibiotics in chronic diarrhoea. In sub-Saharan Africa, chronic diarrhoea often presents with significant weight loss [1,2], eventually occurs in 60–90% of HIV-infected adults [3], and causes significant mortality [4]. Most studies of chronic diarrhoea performed in Africa have enrolled hospitalized rather than ambulatory patients [5,6]. Furthermore, the isolation of few treatable enteric pathogens in most cases of HIV-associated chronic diarrhoea has complicated the development and testing of locally applicable management strategies [7,8]. In this study, we managed cases of chronic diarrhoea presenting to an outpatient facility using a modified version (Fig. 1) [9] of the WHO chronic diarrhoea guideline [10] for HIV-1-infected adults.Fig. 1.: World Health Organization clinical care guidelines for treatment of chronic diarrhoea in HIV-1-infected adults. The ‘Level A’ guideline was modified by the substitution of norfloxacin 400 mg twice a day for trimethoprim–sulfamethoxazole (TMP–SSX) because of more than 64% resistance of enteric bacteria in Kenya to TMP–SSX [9], and the prescription of loperamide at the first visit with severe cases of chronic diarrhoea to decrease the frequency of stools rapidly.A cohort of 380 HIV-1-infected adults was formed over a 6-month period, from July 1997, and was followed every 2 months to January 2000. At the initial visit, CD4 T-lymphocytes were enumerated (Becton Dickenson, Sunnyvale, CA, USA). A modified WHO algorithm was used to manage patients presenting with chronic diarrhoea defined as three or more loose stools intermittently or continuously for at least one month (Fig. 1). Those with a history of antibiotic use in the previous 2 weeks were excluded from participation. The WHO algorithm recommended trimethoprim–sulfamethoxazole (TMP–SSX), metronidazole and loperamide for the management of chronic diarrhoea. Norfloxacin was substituted for TMP–SSX as a result of observed resistance [9]. Loperamide 2 mg was given as an initial dose, followed by repeated 2 mg doses after every episode of diarrhoea if symptoms worsened, remained unchanged, or were particularly severe at initial presentation. Clinical cure was defined as two or less formed stools per day. Stool specimens were obtained for microscopy, culture and sensitivity, including acid-fast staining for Mycobacterium tuberculosis following standard procedures [6]. SPSS for Windows 9.0 (SPSS Inc., Chicago, IL, USA) was used to evaluate potential risk factors for the first episode of chronic diarrhoea and to assess factors related to cure rate. Multivariate logistic regression was performed to calculate adjusted rate ratios (ARR), the final model being selected using backward elimination. As duration of participation varied, length of follow-up was always included as a covariate in the regression. The initial characteristics of this cohort have previously been reported [11]. At enrollment, 29 individuals (8%) had CD4 T-cell counts less than 50 cells/μl, 94 (25%) had counts of 50–199 cells/μl, 159 (42%) had counts of 200–499 cells/μl, and 89 (23%) had counts of 500 cells/μl or greater. According to the CDC classification system for HIV-1 infection, 61 individuals (16%) were asymptomatic, 196 (53%) were symptomatic without an AIDS-defining illness, and 117 (31%) had an AIDS-defining condition [12]. The median length of follow-up was 552 days (range 1–831 days). One hundred and forty-eight episodes of chronic diarrhoea were diagnosed in 85 individuals; 53 suffered from a single episode whereas 33 were diagnosed with more than one episode of chronic diarrhoea. The overall incidence of chronic diarrhoea was 286 per 1000 person-years. In multivariate analysis, while controlling for length of time in the cohort, age 30 years or greater [ARR 1.8, 95% confidence interval (CI) 1.1–2.6], male sex (ARR 1.5, 95% CI 1.0–2.2), initial CD4 T-cell count less than 200 cells/μl (ARR 1.6, 95% CI 1.1–2.2), a history of chronic diarrhea (ARR 2.6, 95% CI 1.5–3.7), and 10% or more weight loss (ARR 1.7, 95% CI 1.1–2.3) during the year before enrollment were associated with an increased risk of chronic diarrhoea. Of the 85 individuals diagnosed with chronic diarrhoea, those with less than 8 years of education [17/33 (52%) in comparison to 8 or more years of education 15/50 (30%) P = 0.05], and not treated with metronidazole for their first episode of chronic diarrhoea [9/15 (60%) versus 23/70 (33%) P = 0.05] more likely suffered recurrent chronic diarrhoea. Overall, 125 out of 148 (85%) episodes of chronic diarrhoea responded completely to treatment. Of the 85 individuals experiencing their first occurrence of chronic diarrhoea, a similar proportion, 72 (85%) also had complete resolution of symptoms as previously defined. Eleven chronic diarrhoea episodes (7%) required intravenous rehydration and hospitalization. Only analysing the first episode of chronic diarrhoea, we constructed a multivariate model to evaluate factors that might be associated with the resolution of symptoms. Not using metronidazole (P < 0.03), the presence of bacteria (not including M. tuberculosis) on culture (P = 0.06), a history at enrollment of chronic diarrhoea (P < 0.03), and being widowed (P < 0.01) were associated with a decreased cure rate. The identification of ova and cysts (P = 0.63) and age (P = 0.10) were not associated with the resolution of chronic diarrhoea, but were retained in the final model as they tended to confound the effect of the other factors. Stool samples were collected for culture in 96 out of 148 episodes of chronic diarrhoea (65%); for the remainder of the cases participants were unable to produce stool at the time of visit. Out of these 96 episodes of chronic diarrhoea, 34 (35%) had potential pathogens detected, of which 22 (62%) were considered to be potentially treatable. Mycobacterium spp. (13%) was the most common organism, followed by Cryptosporidium spp. (11%), Salmonella spp. (6%) and Shigella spp. (3%). Although a few studies have measured the incidence and described risk factors for chronic diarrhoea in HIV-1-infected adults living in developing countries, this is the first attempt to validate a simple treatment algorithm for chronic diarrhoea. Although subjects received virtually all of their medical care at the study clinic, leading to the accurate detection of chronic diarrhoea, loss to follow-up probably caused an underestimation of the true incidence of chronic diarrhoea. The incidence of chronic diarrhoea was less than half that recently found in an HIV-1-infected Ugandan cohort [13]. However, the Ugandan study reported a median CD4 cell count of 215 cells/μl at enrollment in comparison to 300 cells/μl in the current study, and included both acute and chronic cases [13]. Overall, the modified WHO clinical guideline performed well, with 85% of chronic diarrhoea episodes clinically responding to treatment. In Zambia, 43% of HIV-1-infected hospitalized patients with chronic diarrhoea demonstrated spontaneous remission [14], whereas in a separate study [15], 39 and 21%, respectively, showed complete and partial responses after 14 days of treatment with albendazole. The infrequent need to hospitalize and rehydrate patients in our study probably partly accounts for the higher cure rate than found in Zambia. In the future, randomized controlled trials may help explain the role of antibiotics or other newer agents such as nitazoxanide in the management of HIV-1-associated chronic diarrhoea [16].

We evaluated a modified WHO algorithm for the management of chronic diarrhoea in ambulatory HIV-1-infected adults that included the use of norfloxacin, metronidazole and loperamide. In a cohort of 380 HIV-1-infected adults in Nairobi, Kenya, the incidence of chronic diarrhoea was 286 cases/1000 person-years; 85% of chronic diarrhoea episodes responded completely to therapy. Although these findings appear to validate the modified WHO chronic diarrhoea algorithm, randomized controlled trials may better define the indication for antibiotics in chronic diarrhoea.

In sub-Saharan Africa, respiratory tract infections (RTI) are the leading cause of serious morbidity and mortality in HIV-infected persons. This study sought to investigate demographic, socioeconomic, and environmental risk factors for pneumonia in a cohort of HIV-infected women. The authors performed a nested case–control study in a cohort of HIV-1–infected adults followed in Nairobi, Kenya. Thirty-nine women who developed pneumonia during the follow-up period were selected as cases, and 66 women who did not develop pneumonia were randomly chosen to serve as control subjects. A questionnaire was administered in subjects' homes that assessed demographics, home environment, and socioeconomic status. Women were followed in the cohort for a median of 36.8 months (range, 27.3–39.3). Adjusting for length of follow-up period, factors associated with lower socioeconomic status (lower monthly spending [OR = 3.2; 95% CI, 1.2–8.4 per 10,000 Kenyan shilling decrease], having no savings [OR = 4.1; 95% CI, 1.4–11.9], less sturdy home construction material such as mud or cement walls [OR = 2.6; 95% CI, 1.1–5.9] or dirt floors [OR = 2.8; 95% CI, 1.0–7.6], and lack of a window in the home [OR = 5.5; 95% CI, 0.9–32.2]) and being widowed (OR = 4.3; 95% CI, 1.2–15.1) or single (OR = 3.3; 95% CI, 1.0–11.2) were associated with an increased risk of pneumonia. In multivariate analysis, widowed (AOR = 5.9; 95% CI, 1.3–26.3), single (AOR = 7.7; 95% CI, 1.6–36.4), and divorced (AOR = 4.5; 95% CI, 1.0–20.1) women, those without savings (AOR = 3.7; 95% CI, 1.2–11.7), and those living in more crowded and contagious conditions (AOR = 1.5; 95% CI, 1.1–2.1) remained at increased risk of pneumonia. If confirmed by prospective investigation, these findings could help identify persons and subpopulations of HIV-infected women with the greatest risk of pneumonia.

To evaluate the WHO (World Health Organization) algorithm for management of respiratory tract infection (RTI) in HIV-1-infected adults and determine risk factors associated with RTI, we enrolled a cohort of 380 HIV-1-seropositive adults prospectively followed for incident RTI at an outpatient clinic in Nairobi, Kenya. RTI was diagnosed when patients presented with history of worsening or persistent cough. Patients were treated with ampicillin, or antituberculosis therapy when clinically indicated, as first-line therapy and with trimethoprim/sulfamethoxazole as second-line therapy. Five hundred ninety-seven episodes of RTI were diagnosed: 177 of pneumonia and 420 of bronchitis. The WHO RTI algorithm was used for 401 (95%) episodes of bronchitis and 151 (85%) episodes of pneumonia (p < .001). Three percent of bronchitis cases versus 32% of pneumonia cases failed to respond to first- or second-line treatment (p < .0001). Being widowed (adjusted odds ratio [OR] = 2.1, 95% confidence interval (CI: 1.0-4.4), less than 8 years of education (adjusted OR = 2.5. CL 1.5-4.1), and CD4 count < 200 cells/mul (adjusted OR = 2.4, Cl: 1.4-3.9) were risk factors for pneumonia. A high percentage of patients (32%) with pneumonia required a change in treatment from that recommended by the WHO guidelines. Randomized trials should be performed to determine more appropriate treatment strategies in HIV-1-infected individuals.

HIV infection has now been consistently identified as the major cause of death in young Africans in both urban and rural areas. In Africa, several studies have defined the clinical presentation of HIV disease but there have only been a limited number of autopsy studies. Because of the scarcity of autopsy data and the possibility of differing type and frequency of opportunistic infections between different geographic locations we set out to study consecutive new adult medical admissions to a tertiary referral hospital in Nairobi and perform autopsies on a sample of HIV-1-positive and HIV-1-negative patients who died in the hospital ward. Basic demographic data were collected on all patients admitted to two acute medical wards over an 11-month period. Final outcome and final clinical diagnoses were recorded at discharge or death. An autopsy examination was requested if the patient died in the ward. Autopsy examination was performed in 75 HIV-1-positive (40 men, 35 women) and 47 HIV-1-negative (28 men, 19 women) adults who died in the hospital. This represented 48.4% of all HIV-1-positive deaths and 33.3% of all HIV-1-negative deaths. Tuberculosis (TB) and bacterial and interstitial bronchopneumonia accounted for 96% of the major pathology in patients found to be HIV-1-positive at autopsy. TB was present in half the HIV-1-positive autopsy patients and was disseminated in over 80% of cases. Meningeal involvement was present in 26% of those with disseminated TB. By contrast, TB was much less common in the HIV-1-negative patients at autopsy in whom bacterial bronchopneumonia and malignancies were the most common pathologies. The type pathology found in the HIV-1-positive autopsy patients was not different than that found in other areas in Africa so far studied.

Objectives: Chronic diarrhoea and wasting are well recognized features of AIDS in Africa, However, because of resource constraints few comprehensive aetiological studies have been conducted in sub-Saharan Africa which have included a broad range of microbiological investigations. We undertook a prospective cross-sectional study of adult patients admitted to a government hospital in Nairobi, Kenya, to determine possible bacterial, mycobacterial, parasitic and viral causes of diarrhoea; to consider which may be treatable; and to relate microbiological findings to clinical outcome.Methods: Stool specimens from 75 consecutive HIV-seropositive patients with chronic diarrhoea admitted to a Nairobi hospital were subjected to microbiological investigation and results were compared with clinical findings and outcome. Stool samples were cultured for bacteria and mycobacteria and underwent light and electron microscopy; lawns of Escherichia coli were probed for pathogenic types and aliquots were tested for the presence of Clostridium difficile cytotoxin. Blood cultures for mycobacteria and other bacterial pathogens were performed as clinically indicated.Results: Thirty-nine (52%) patients yielded putative pathogens, the most common being Cryptosporidium sp. (17%), Salmonella typhimurium (13%), and Mycobacterium tuberculosis (13%). Of 41 patients investigated for pathogenic Escherichia coli, enteroaggregative E. coli and diffusely adherent B. coli were each found in four patients. Thirty-one (41%) patients died. Detection of cryptosporidium cysts was the single most significant predictor of death (X-2 = 5.2, P < 0.05). Many patients did not improve (21; 28%) or self-discharged whilst still sick (5; 7%) but five (7%) were diagnosed ante mortem with tuberculosis and treated and a further 13 (17%) showed improvement by time of discharge.Conclusions: HIV-infected patients with chronic diarrhoea in Nairobi have a poor outcome overall, and even with extensive investigation a putative pathogen was identified in only just over half the patients. The most important step is to exclude tuberculosis; and the most useful investigation appears to be Ziehl-Neelsen staining. Other potentially treatable Gram-negative bacterial pathogens, S. typhimurium, Shigella sp. and adherent E. coli were, however, common but require culture facilities which are not widely accessible for definitive identification. Further studies focussing on simple ways to identify sub-groups of patients with treatable infections are warranted.

Although significant bacteriuria and urinary tract infection are more common in immunocompetent women than men, studies linking HIV immunosuppression with an increased risk of developing urinary infection have so far only been carried out in men. We therefore examined the relationship between bacteriuria and HIV status and CD4+cell count in a relatively homogeneous cohort of female commercial sex workers (CSW) attending a community clinic in Nairobi. Two hundred and twenty-two women were enrolled, and grouped according to HIV status and CD4 count. Group 1 were HIV seronegative (n = 52); Group 2 were HIV seropositive with CD4 + counts above 500 x 10(6)/l (n = 51); Group 3 were HIV seropositive with CD4 + counts between 201 and 500 x 10(6)/l (n = 67); Group 4 were HIV seropositive with CD4+counts below 200 x 10(6)/l (n = 52). Clinical signs and symptoms were noted and mid-stream specimens of urine obtained for culture and sensitivity. Overall 23% (50/222) had significant bacteriuria. The rates in each group respectively were 25%, 29%, 19% and 23% and there was no significant association between bacteriuria and HIV status; or between bacteriuria and level of immuno-suppression as indicated by CD4 + count. Overall 19% (30/222) of women had symptoms (frequency; dysuria; loin pain; smelly urine) or signs (fever; loin tenderness) compatible with urinary tract infection. However there was no significant association between symptoms or signs of infection and bacteriuria or HIV status. A typical range of pathogens, predominantly Enterobacteriaceae, were isolated and there were high rates of resistance to commonly used antimicrobials as well as 10% resistance to ciprofloxacin. Although high rates of significant bacteriuria can occur in highly sexually-active women, this appears unrelated to HIV infection or the level of HIV-related immunosuppression and is generally asymptomatic or clinically indistinct.

Previous studies from Africa have been unable to identify disseminated Mycobacterium avium complex (MAC) infection in patients with advanced human immunodeficiency virus (HIV) infection. We performed mycobacterial blood cultures and CD4 counts on 48 symptomatic adults with advanced HIV infection admitted to the hospital in Nairobi, Kenya over 4 weeks in 1992. Fourteen patients had mycobacteremia; these patients had significantly lower CD4 counts than the patients with negative cultures (14/mm3 vs. 85/mm3; p < 0.01). Three patients (6%) were bacteremic with M. avium (mean CD4 count, 10/mm3) and 11 (23%) were bacteremic with Mycobacterium tuberculosis complex (MTB) (mean CD4 count, 15/mm3). Thus, M. avium bacteremia was detected significantly less frequently in the study population than MTB bacteremia (p = 0.04). The minimum rate for HIV-associated disseminated M. avium infection in patients admitted to the hospital in Nairobi was estimated to be approximately 1%. Patients with mycobacteremia died or were discharged home sick before the diagnosis was made. Disseminated M. avium does occur in adults with advanced HIV infection in sub-Saharan Africa, but is less common than disseminated MTB.