Search

We developed, implemented, and evaluated a myocardial infarction (MI) adjudication protocol for cohort research of human immunodeficiency virus. Potential events were identified through the centralized Centers for AIDS Research Network of Integrated Clinical Systems data repository using MI diagnoses and/or cardiac enzyme laboratory results (1995-2012). Sites assembled de-identified packets, including physician notes and results from electrocardiograms, procedures, and laboratory tests. Information pertaining to the specific antiretroviral medications used was redacted for blinded review. Two experts reviewed each packet, and a third review was conducted if discrepancies occurred. Reviewers categorized probable/definite MIs as primary or secondary and identified secondary causes of MIs. The positive predictive value and sensitivity for each identification/ascertainment method were calculated. Of the 1,119 potential events that were adjudicated, 294 (26%) were definite/probable MIs. Almost as many secondary (48%) as primary (52%) MIs occurred, often as the result of sepsis or cocaine use. Of the patients with adjudicated definite/probable MIs, 78% had elevated troponin concentrations (positive predictive value = 57%, 95% confidence interval: 52, 62); however, only 44% had clinical diagnoses of MI (positive predictive value = 45%, 95% confidence interval: 39, 51). We found that central adjudication is crucial and that clinical diagnoses alone are insufficient for ascertainment of MI. Over half of the events ultimately determined to be MIs were not identified by clinical diagnoses. Adjudication protocols used in traditional cardiovascular disease cohorts facilitate cross-cohort comparisons but do not address issues such as identifying secondary MIs that may be common in persons with human immunodeficiency virus.

Anemia, a common hematologic complication in human immunodeficiency virus (HIV)-infected patients, can be caused by mechanisms including infections, neoplasms, or drug treatment. Studies have consistently found anemia to be associated with reduced survival, even when potentially confounding factors were controlled for. Importantly, recovery from anemia has been shown to reduce this risk to approximately the same level as seen among patients never having had anemia. Although anemia traditionally has been treated with blood transfusions, recent studies have shown recombinant human erythropoietin (r-HuEPO) to be effective in elevating hematocrit values and reducing transfusion requirements in HIV-infected patients who have endogenous erythropoietin levels of < or = 500 IU/L. Therapy with r-HuEPO has been shown to be safe and well tolerated. In a recent study, moreover, receipt of erythropoietin was associated with a decreased risk of death, whereas transfusion was associated with an increased risk. If these results are confirmed, the link between r-HuEPO and decreased risk of death in HIV-infected patients with anemia will be further strengthened.

The effect of acute coinfections on plasma human immunodeficiency virus (HIV) load and immune activation markers was evaluated. Thirty-two HIV-infected persons were prospectively enrolled; 18 had pre-illness, acute, and follow-up specimens. Plasma HIV RNA levels were determined by reverse transcriptase-polymerase chain reaction, and serum levels of activation markers, including tumor necrosis factor (TNF)-alpha, soluble (s) TNF receptors (R)-I and -II, interleukin (IL)-2, IL-6, IL-10, sIL-2R, sCD4, and sCD8, were assessed by commercial ELISAs. Median plasma HIV load increased 7. 8-fold during illness (P=.001) and decreased 1.5-fold (P=.01) during convalescence (median, 15 days). Significant virus load reductions were limited to subjects with clinical recovery. By regression analysis, changes in plasma HIV RNA were significantly associated with changes in sTNFR-I, sTNFR-II, and sIL-2R. Increased HIV replication during acute coinfections is associated with in vivo immune activation, which underscores the need to prevent and promptly treat intercurrent illnesses.

Although 10%-15% of patients with AIDS in the United States may acquire cryptosporidium infection, little data exist on clinical or histological characteristics that differentiate clinical outcomes. A case-control study of 83 HIV-positive adult patients with cryptosporidiosis was conducted, as was a histopathologic review of data on gastrointestinal biopsy specimens from 30 patients. Four clinical syndromes were identified: chronic diarrhea (36% of patients), choleralike disease (33%), transient diarrhea (15%), and relapsing illness (15%). A multivariate analysis of data for cases and controls revealed that acquiring cryptosporidiosis was associated with the presence of candidal esophagitis (odds ratio [OR], 2.53; P < .002) and Caucasian race (OR, 6.71; P = .0001) but not with sexual orientation. Cases had a significantly shorter duration of survival from the time of diagnosis than did controls (240 vs. 666 days, respectively; P = .0004), which was independent of sex, race, or or injection drug use. Antiretroviral use was protective against disease (OR, 0.072; P = .0001). All four clinical syndromes were represented among the histological data. There was no statistically significant correlation between histological intensity of infection and clinical severity of illness.

Vancomycin-resistant Enterococcus (VRE) is a major nosocomial pathogen. We collected clinical and laboratory data on 93 hospitalized adults with VRE bacteremia and 101 adults with vancomycin-susceptible enterococcal (VSE) bacteremia. Risk factors for VRE bacteremia included central venous catheterization, hyperalimentation, and prolonged hospitalization prior to the initial blood culture. VRE-infected patients were less likely to have undergone recent surgery or have polymicrobial bacteremia, suggesting a pathogenesis distinct from traditional VSE bacteremia. Prior exposure to metronidazole was the only significant pharmacologic risk factor for VRE bacteremia. Animal studies suggest metronidazole potentiates enterococcal overgrowth in the gastrointestinal tract and translocation into the bloodstream. An increasing APACHE II score was the major risk factor for death in a multivariate analysis, with VRE status being of only borderline significance.

To evaluate the impact of Cryptosporidium infection on diarrheal disease burden and nutrition status, a nested case-control study was done among children who were followed from birth in Fortaleza, Brazil. The diarrhea history and growth records of 43 children with a symptomatic diarrhea episode of cryptosporidiosis (case-children) were compared with those of 43 age-matched controls with no history of cryptosporidiosis. After Cryptosporidium infection, case-children < or = 1 year old experienced an excessive and protracted (nearly 2 years) diarrheal disease burden. Case-children < or = 1 year old with no history of diarrhea prior to their Cryptosporidium infection also experienced a subsequent increased diarrheal disease burden with an associated decline in growth. Control subjects experienced no change in their diarrhea burden over time. This study suggests that an episode of symptomatic Cryptosporidium infection in children < or = 1 year of age is a marker for increased diarrhea morbidity.

A cross-sectional study was conducted to assess the prevalence and microbiology of oral infection due to fluconazole-resistant Candida in patients with AIDS. Oral swab specimens for fungal cultures were obtained from 100 consecutive outpatients with CD4 lymphocyte counts of < 200/mm3. At least one fungal organism demonstrating in vitro resistance to fluconazole (minimum inhibitory concentration, > or = 8 micrograms/mL) was isolated from 26 (41%) of 64 patients for whom cultures were positive. When fluconazole-resistant C. albicans was isolated, in vitro resistance correlated with clinical thrush. None of 10 patients from whom only non-albicans species of Candida were isolated had active thrush. The patients from whom fluconazole-resistant Candida albicans was isolated had lower CD4 cell counts (median, 9/mm3), a greater number of treated episodes of thrush (median, 4.5), and a greater median duration of prior fluconazole treatment (231 days) than did patients from whom fluconazole-susceptible C. albicans was isolated (median CD4 cell count, 58/mm3 [P = .004]; median number of treated episodes of thrush, 2.0 [P = .001]; and median duration of prior fluconazole treatment, 10 days [P = .01]; respectively). In a multivariate analysis, the number of episodes and duration of fluconazole therapy were independent predictors of resistance.

To identify risk factors for pneumococcal infection among human immunodeficiency virus-infected patients, a nested case-control study was done in an urban university human immunodeficiency virus clinic. Subjects with pneumococcal illness seen between 1 January 1990 and 1 July 1994 (n=85) were randomly matched to controls from the same population. Patients with pneumococcal disease were more likely than controls to be African Americans (adjusted odds ratio [OR]=3.92), have <200 CD4 cells/mm3 (adjusted OR=3.38), have a history of any pneumonia (adjusted OR=3.28), and have an albumin level of <3.0 g/dL (adjusted OR=6.25). Use of zidovudine (adjusted OR=0.38) and pneumococcal vaccination when the subject had >200 CD4 cells/mm3 (adjusted OR=0.22) were less common in cases than in controls. Similar results were found when only cases with infections of usually sterile sites were analyzed. Pneumococcal vaccine may be most protective when it is administered before advanced immunodeficiency develops.

In many patients, the etiology of community-acquired pneumonia is not known but may be caused by previously undescribed pathogens in some cases. The recently identified hantavirus Sin Nombre (SN) causes hantavirus pulmonary syndrome. Because sporadic cases have occurred outside the range of its reservoir (the deer mouse Peromyscus maniculatus), an investigation sought to determine whether hantaviruses contributed to cases of community-acquired pneumonia in a large Baltimore hospital. Acute-phase sera from 385 hospitalized patients with pneumonia were examined using an IgG ELISA technique with antigens prepared from several hantaviruses: prototype Hantaan (HTN), Seoul (SEO), Puumala (PUU), Convict Creek (HN107), and SN. Of 385 sera, 8 (2.1%) showed some reactivity with one or more HTN, SEO, or PUU antigens but none had detectable specific IgM antibodies. No sera were reactive with SN or HN107 antigens. Thus, hantaviruses are an uncommon cause of community-acquired pneumonia in the Baltimore area.

In a case-control study to identify risk factors for fluconazole-resistant oroesophageal candidiasis in human immunodeficiency virus-infected patients, 25 patients with clinical and in vitro fluconazole-resistant candidiasis were paired with controls who had treatment-responsive candidiasis and who had been observed for similar time periods. After their first episode of candidiasis, patients who later developed fluconazole resistance had more treated episodes than did matched controls (cases, 3.1; controls, 1.8; P = .004), lower median CD4 cell counts (11/mm3 vs. 71/mm/3; P = .004), and greater median durations of all antifungal therapy (419 vs. 118 days; P < .001) and of systemic azole therapy (272 vs. 14 days; P < .001). When paired with a second set of controls matched by CD4 cell count as well as first diagnosis of candidiasis, cases continued to show greater median exposure to azoles (272 vs. 88 days; P = .005). These data indicate that advanced immunosuppression and exposure to oral azoles are risk factors for the development of fluconazole resistance.

To determine the incidence and clinical manifestations of herpes zoster in a hospital-based clinic for adults infected with human immunodeficiency virus (HIV), we reviewed the records of all patients for whom zoster was diagnosed at or after their first clinic visit. Fifty-two episodes of zoster occurred in 45 patients during 1,614 person-years of follow-up (incidence, 3.2 episodes per 100 person-years). The following major complications of zoster occurred in 12 patients (27%): ocular complications (5), neurological complications (4), and chronic atypical skin lesions (5). Six patients each had postherpetic neuralgia and bacterial superinfection, which were the common minor complications of zoster. Multivariate analysis revealed that only a low CD4 cell count (< or = 200/mm3) was predictive of a major complication of zoster (OR, 13.2; 95% CI, 1.52-114; P = .019). Thus, complications of herpes zoster are common in patients with HIV infection, especially those with advanced immunosuppression.

To evaluate the association between acyclovir use and survival in patients with advanced human immunodeficiency virus infection, observational data from 1044 persons with AIDS or AIDS-related complex (ARC) and < or = 250 CD4 cells/mm3 following initiation of zidovudine were analyzed. Of these patients, 336 (32%) received regular acyclovir (> or = 6 weeks in 2 months). There were no differences in mortality data between acyclovir users and nonusers overall or when analyzed from 1 year after first use of zidovudine, from time of AIDS in those with ARC at enrollment, from patients with AIDS or < 100 CD4 cells/mm3 at enrollment, or from patients taking acyclovir for up to 10 months. Acyclovir use was associated with increased mortality (relative hazard, 1.28; P = .057) independent of herpesvirus infections and of other characteristics associated with mortality. In this study, the use of acyclovir at doses for treatment of herpes simplex virus infection in combination with zidovudine was not associated with prolonged survival.

To determine the prevalence, incidence, and effects on disease progression and survival of herpes zoster in patients with advanced human immunodeficiency virus (HIV) disease, data from a multicenter observational cohort study of 1044 patients with AIDS or AIDS-related complex (ARC) and CD4 cell count < or = 0.25 x 10(9)/L treated with zidovudine were analyzed. Of 163 patients (16%) with a history of herpes zoster at enrollment, 22 (13%) had a recurrence during the 2-year follow-up. For those without prior herpes zoster, the probability of its development was 6.3% at 1 and 8.8% at 2 years. Progression to AIDS was not associated with herpes zoster. By proportional hazards analysis, an initial occurrence of herpes zoster was associated with prolonged survival independent of baseline CD4 cell count and disease stage; however, recurrence tended to be associated with death. Thus, herpes zoster is relatively common in advanced HIV infection and its initial occurrence late in disease may indicate improved prognosis.

Data were analyzed from a multicenter observational cohort study of 1002 persons with AIDS or AIDS-related complex (ARC) and total CD4 cell count < 0.25 x 10(9)/L treated with zidovudine between April 1987 and April 1988. Cytomegalovirus (CMV) disease developed in 109 patients (10.9%), with a 2-year actuarial risk of 15%. Manifestations included retinitis (93 patients), esophagitis (10), colitis (8), gastritis (1), hepatitis (1), and encephalitis (1). The probability of CMV disease at 2 years for patients with initial counts < 0.1 x 10(9)/L was 21.4%, compared with 10.3% for patients with initial counts > or = 0.1 x 10(9)/L (P < .001). By proportional hazards analysis, baseline CD4 cell count < 0.1 x 10(9)/L, enrollment diagnosis of AIDS, and homosexuality were significantly associated with subsequently developing CMV disease. Median survival after diagnosis of CMV disease was 173 days, and CMV was an independent predictor of death. CMV contributes to AIDS-related morbidity and mortality. As new anti-CMV drugs become available, prophylaxis should be targeted at individuals with CD4 cell counts < 0.1 x 10(9)/L.

In a previous report (Kimball and Friedman, Am J Epidemiol, 1992;135:1279-86), linear models for relating health outcomes to alcohol consumption were proposed for differentiating between beverage type effects and beverage preference effects. The models were applied to data relating serum high density lipoprotein cholesterol to alcohol consumption. In this report, those models are extended to the nonlinear case and are applied to data from the 1982 Maryland Hypertension Survey relating systolic blood pressure to alcohol consumption.

A randomized double-blind clinical trial was done of aztreonam versus aminoglycoside therapy for the empiric treatment of seriously ill nonneutropenic patients suspected of aerobic gram-negative bacterial infection. Each patient was treated for greater than or equal to 72 h with the study drug. Nephrotoxicity, defined by greater than or equal to 50% increase in baseline serum creatinine, occurred in 12 (15%) of 92 patients receiving aminoglycoside therapy and 1 (1%) of 92 patients receiving aztreonam (P less than .004). More severe nephrotoxicity, defined by greater than or equal to 100% increase in baseline serum creatinine, occurred in 6 (6.5%) of 92 patients receiving aminoglycoside therapy and in 1 of 92 receiving aztreonam (P less than .11). Patients with an elevated baseline total bilirubin level were most likely to develop nephrotoxicity. Auditory toxicity occurred in 2 (7%) of 28 evaluatable patients receiving aminoglycoside therapy and in 1 (3%) of 33 receiving aztreonam (P less than .58). One patient, who received aminoglycoside, developed vestibular toxicity. In nonneutropenic patients believed to be at increased risk for renal dysfunction, aztreonam is a less toxic alternative to aminoglycoside therapy for treatment of suspected aerobic gram-negative infection.

Whether ceftazidime monotherapy is equal in efficacy to combination regimens (Comb) for empiric treatment of febrile neutropenic patients was tested using meta-analysis. Published studies and abstracts of ceftazidime trials were identified, their quality assessed, the efficacy data abstracted and pooled, and effects of patient and study characteristics examined. The pooled odds ratio (OR) of failure of ceftazidime for febrile episodes was 1.27 (95% confidence interval [CI]: 0.79-2.03; n = 1077) and for bacteremic episodes was 0.72 (CI, 0.33-1.58; n = 248; OR less than 1.0 favors ceftazidime). Results were not significantly affected by type of antibiotic in Comb, age, neutropenia (less than 500/mm3), study quality, or combining abstracts. Results indicate that Comb does not offer a significant advantage over ceftazidime. A subgroup of profoundly neutropenic (less than 100/mm3) patients could not be assessed, raising the possibility that in this important subgroup monotherapy and Comb may not be equivalent. Use of ceftazidime empirically may require modification based on microbiologic results and clinical course.

This study examines the relationship between alcohol consumption and blood pressure in the 1982 Maryland Hypertension Survey, a crossectional population-based household survey of blood pressure control in adults residing in Maryland. In individuals less than 50 years old, a J shaped dose-response association was found with abstainers and heavy alcohol consumers having significantly higher blood pressures than moderate alcohol consumers (1 to 2 beverages per day). In individuals 50 years and older, alcohol was associated with higher blood pressures only at the highest levels of intake (greater than 2 beverages per day). The prevalence of hypertension was similarly affected in each age group. This association between alcohol consumption and blood pressure was independent of several variables that are associated with increased blood pressure such as age, sex, race, smoking, education, Quetelet index, social participation, and physical activity. The population attributable risk for hypertension due to heavy alcohol consumption is 5 to 7% in those greater than 50 years old and 6 to 8% in those less than 50 years old. These data suggest that alcohol consumption is a potentially important risk factor for elevations in blood pressure and hypertension.

Alcohol use is a recognized risk factor for stroke. We hypothesized that alcohol use may increase the risk of stroke, independent of an effect on blood pressure, by causing cerebral vasoconstriction. To examine this, we used retinal vessels as a marker for cerebral vessels and analysed the cross-sectional associations between alcohol use and total retinal vessel width in 741 Japanese and 434 American white male telephone executives. Systolic and diastolic blood pressures were negatively associated with retinal vessel width in the Japanese (p less than 0.0001, 0.0001, respectively) but this association did not achieve statistical significance in the Americans (p less than 0.1, 0.3). Japanese drinkers had a larger mean retinal vessel width than Japanese abstainers, while American drinkers had a smaller retinal vessel width than abstainers. These associations between alcohol use and retinal vessel width were not significant within nationalities but the interaction between alcohol use and nationality was significant, independent of other stroke risk factors (p less than 0.04). Although these results do not support the initial hypothesis, they are consistent with known biological differences in alcohol metabolism between Japanese and American men and support different effects of alcohol use on retinal vessel calibre in these two nationalities.

Risk factors for the development of auditory toxicity in patients receiving aminoglycosides were determined from the analysis of 135 patients enrolled in three prospective, randomized, double-blind clinical trials of gentamicin, tobramycin, and amikacin. Auditory toxicity, defined as a decrease in auditory acuity of greater than or equal to 15 dB, occurred in 30 patients (22.3%). Patients with auditory toxicity underwent therapy for a longer period, were more likely to be bacteremic, and had, on the average, a higher temperature (P less than 0.05). Using stepwise discriminant analysis, we selected these factors with liver dysfunction and the serum urea nitrogen:serum creatinine ratio in a function that accurately discriminates between toxic and nontoxic patients. Factors not adding significantly to the predictive accuracy of the equation were plasma aminoglycoside levels, aminoglycoside type, furosemide use, diabetes, age, sex, renal function, initial auditory acuity, hematocrit value, and shock. This analysis may be important both for determining the pathophysiology of auditory toxicity and for the prognostic stratification of patients receiving aminoglycosides in clinical trials.

To determine the association of aminoglycoside levels with mortality from gram-negative bacteremia, we analyzed the case reports of patients from four prospective, randomized, and controlled clinical trials of gentamicin, tobramycin, and amikacin. Twelve (13.5%) of 89 patients died. One (2.4%) death occurred in 41 patients with early (1-hr postinfusion) peak concentrations of greater than 5 micrograms/ml of gentamicin and tobramycin and of greater than 20 micrograms of amikacin/ml; nine deaths (20.9%) occurred in 43 patients with lower concentrations. Five (8.3%) deaths occurred in 60 patients with mean peak concentrations for the entire course of therapy of greater than 5 micrograms/ml of gentamicin and tobramycin and of greater than 20 micrograms of amikacin/ml; five (20.8%) deaths occurred in 24 patients with lower concentrations. Stepwise discriminant analysis showed that therapeutic early peak concentration was a significant factor in the presence of three other factors: severity of underlying illness, peak temperature, and initial leukocyte count. The results suggest the importance of achieving adequate early aminoglycoside levels in patients with gram-negative bacteremia.

We have previously identified risk factors for nephrotoxicity in 204 patients receiving aminoglycosides. We added data on an additional 134 patients from the University of Chicago who were treated for longer periods (13.1 +/- 9.4 days) to determine if duration of therapy was also a risk factor. In the combined populations, nephrotoxicity developed in 59 (17.5%) of 338 patients, and duration of therapy was the factor most strongly associated with nephrotoxicity (P less than .0001). A new predictive model was developed that was then tested prospectively on 175 consecutive medical and surgical patients who had plasma levels of aminoglycoside measured. The new model accurately identified 14 of 15 patients with nephrotoxicity (sensitivity = 93%) and 106 of 160 without nephrotoxicity (specificity = 66%). A bedside scoring system accurately identified high-risk and low-risk patients. We believe these data add to the understanding of the clinical correlates of aminoglycoside nephrotoxicity and aid in the identification of high-risk patients.

In an examination of the relationships among plasma aminoglycoside concentrations, the minimal inhibitory concentration (MIC) for the infecting organism, and therapeutic outcome, data were analyzed from 236 patients with gram-negative bacterial infections who were participants in four clinical trials of gentamicin, tobramycin, and amikacin. Clinical response to therapy occurred in 188 (80%) patients. Elevated maximal and mean peak aminoglycoside concentration/MIC ratios were strongly associated with clinical response (P less than .00001 and P less than .0001, respectively). A graded dose-response effect was found between an increasing maximal peak concentration/MIC ratio and clinical response. By logistic regression the peak concentration/MIC ratios were associated significantly with clinical response after adjustment for underlying severity of illness and other factors correlated with response. These results demonstrate that a high peak concentration relative to the MIC for the infecting organism is a major determinant of the clinical response to aminoglycoside therapy.