Peter A. Selwyn

A Prospective Study of the Risk of Tuberculosis among Intravenous Drug Users with Human Immunodeficiency Virus Infection

To determine the risk of active tuberculosis associated with human immunodeficiency virus (HIV) infection, we prospectively studied 520 intravenous drug users enrolled in a methadone-maintenance program. Tuberculin skin testing and testing for HIV antibody were performed in all subjects. Forty-nine of 217 HIV-seropositive subjects (23 percent) and 62 of 303 HIV-seronegative subjects (20 percent) had a positive response to skin testing with purified protein derivative (PPD) tuberculin before entry into the study. The rates of conversion from a negative to a positive PPD test were similar for seropositive subjects (15 of 131; 11 percent) and seronegative subjects (26 of 202; 13 percent) who were retested during the follow-up period (mean, 22 months). Active tuberculosis developed in eight of the HIV-seropositive subjects (4 percent) and none of the seronegative subjects during the study period (P less than 0.002). Seven of the eight cases of tuberculosis occurred in HIV-seropositive subjects with a prior positive PPD test (7.9 cases per 100 person-years, as compared with 0.3 case per 100 person-years among seropositive subjects without a prior positive PPD test; rate ratio, 24.0; P less than 0.0001). We conclude that, although the prevalence and incidence of tuberculous infection were similar for both HIV-seropositive and HIV-seronegative intravenous drug users, the risk of active tuberculosis was elevated only for seropositive subjects. These data also suggest that in HIV-infected persons tuberculosis most often results from the reactivation of latent tuberculous infection; our results lend support to recommendations for the aggressive use of chemoprophylaxis against tuberculosis in patients with HIV infection and a positive PPD test.

Clinical Manifestations and Predictors of Disease Progression in Drug Users with Human Immunodeficiency Virus Infection

BACKGROUND AND METHODS To examine the clinical course of human immunodeficiency virus (HIV) infection in injection-drug users, we conducted a prospective study of a cohort of patients in a methadone-treatment program in New York City from July 1985 through December 1990. The patients underwent standardized evaluations at base line and semiannually thereafter and received on-site primary medical care. Rates of progression to the acquired immunodeficiency syndrome (AIDS) and major outcomes before the development of AIDS were examined by univariate analyses; the risk of AIDS was also assessed by product-limit survival analysis and proportional-hazards regression. RESULTS Of 318 HIV-seropositive patients who did not yet have AIDS (171 men and 147 women), 90 were black, 179 were Hispanic, and 49 were white; the median age was 33 years. Over a median of 3.0 years of follow-up, 55 (17 percent) received a diagnosis of AIDS (incidence per 100 person-years, 5.8). Major outcomes before the development of AIDS included oral candidiasis (incidence per 100 person-years, 11.2), pyogenic bacterial infections including pneumonia and sepsis (8.0), pulmonary tuberculosis (1.2), multiple constitutional symptoms (13.6), and herpes zoster (1.3). There were 41 deaths from AIDS, and 13 seropositive patients without AIDS (4.1 percent) died of bacterial infections, as compared with only 1 of 411 seronegative patients studied (P < 0.001). The incidence of AIDS was 62 percent lower among those who took zidovudine than among those who did not (P = 0.02). In the multivariate analysis, progression to AIDS was best predicted by low numbers and percentages of CD4+ lymphocytes, nonuse of zidovudine, and the presence of oral candidiasis, bacterial infections, or tuberculosis. There was no consistent relation between progression to disease and the continued use of injection drugs. CONCLUSIONS HIV-infected injection-drug users have progression to AIDS at rates comparable to those of other HIV-infected groups, but they have substantial pre-AIDS morbidity and mortality, particularly from bacterial infections, which also appear to predict disease progression.

Increased risk of bacterial pneumonia in HIV-infected intravenous drug users without AIDS.

Although patients with AIDS have been noted to be at risk for bacterial pneumonia as well as opportunistic infections, little is known about the risk of bacterial pneumonia in HIV-infected populations without AIDS. To determine the incidence of bacterial pneumonia in a well defined population of intravenous drug users (IVDUs), and to examine any association with HIV infection, we prospectively studied 433 IVDUs without AIDS, enrolled in a longitudinal study of HIV infection in an out-patient methadone maintenance program. At enrollment, 144 (33.3%) subjects were HIV-seropositive, 289 (66.7%) were seronegative. Over a 12-month period, 14 out of 144 (9.7%) seropositive subjects were hospitalized for community-acquired bacterial pneumonia, compared with six out of 289 (2.1%) seronegative subjects. The cumulative yearly incidence of bacterial pneumonia was 97 out of 1000 for seropositives and 21 out of 1000 for seronegatives (risk ratio = 4.7, P less than 0.001). Eleven out of 14 (78.6%) cases among the seropositive patients were due to either Streptococcus pneumoniae [5] or Hemophilus influenzae [6]. Two out of 14 (14.3%) cases among the seropositives were fatal. Stratifying by level of intravenous drug use indicated that even among subjects not reporting active intravenous drug use at study entry, eight out of 82 (9.8%) seropositives compared with three out of 211 (1.4%) seronegatives were hospitalized for bacterial pneumonia over the study period (risk ratio = 6.9, P less than 0.01). This study shows a markedly increased incidence of bacterial pneumonia associated with HIV infection in IVDUs without AIDS.(ABSTRACT TRUNCATED AT 250 WORDS)

Acceptance and adherence with antiretroviral therapy among HIV-infected women in a correctional facility

We examined attitudinal and demographic correlates of antiretroviral acceptance and adherence among incarcerated HIV-infected women. Structured interviews were conducted with 102 HIV-infected female prisoners eligible for antiretroviral therapy. Three quarters of the women were currently taking antiretroviral agents, of whom 62% were adherent to therapy. Satisfaction was very high with the HIV care offered at the prison; 67% had been first offered antiretroviral agents while in prison. Univariate and multivariate analyses showed acceptance of the first offer of antiretroviral therapy to be associated with trust in medication safety, lower educational level, and non-black race. Current acceptance of therapy was associated with trust in the medications efficacy and safety. Medication adherence was correlated with the patient-physician relationship and presence of emotional supports. Nearly one half of these HIV-seropositive women were willing to take experimental HIV medications in prison. This was correlated with satisfaction with existing health care, the presence of HIV-related social supports, and perceived susceptibility to a worsening condition. Acceptance and adherence with antiretroviral agents appear to be significantly associated with trust in medications, trust in the health care system, and interpersonal relationships with physicians and peers. Development of models of care that encourage and support such relationships is essential for improving adherence to antiretroviral therapy, especially for populations that have historically been marginalized from mainstream medical care systems.

Heroin use during methadone maintenance treatment: the importance of methadone dose and cocaine use.

OBJECTIVES The purpose of this study was to examine factors associated with heroin use during methadone maintenance treatment. METHODS Logistic regression statistical models were used to examine data obtained in a cross-sectional sample of 652 methadone patients. RESULTS Heroin use during the 3 months prior to interview was shown to be greatest among (1) patients maintained on methadone dosages of less than 70 mg/day (adjusted odds ratio [OR] = 2.1, 95% confidence interval [CI] = 1.3, 3.4) and (2) patients who used cocaine during treatment (adjusted OR = 5.9, 95% CI = 3.8, 9.1). These results were independent of treatment duration, treatment compliance, alcohol use, and socioeconomic factors. Cocaine users were more likely than nonusers of cocaine to use heroin at all methadone dosage levels. CONCLUSIONS This study confirms and extends past research showing high-dose methadone maintenance to be important to heroin abstinence. Further investigation of the independent association between heroin use and cocaine use is needed.

Preliminary report of the integration of a palliative care team into an intensive care unit

Nearly half of Americans who die in hospitals spend time in the intensive care unit (ICU) in the last 3 days of life. Minority patients who die in the ICU are less likely to formalize advance directives and surviving family members report lower satisfaction with the provision of information and sensitivity to their cultural traditions at the end-of-life. This is a descriptive report of a convenience sample of 157 consecutive patients served by a palliative care team which was integrated into the operations of an ICU at Montefiore Medical Center in the Bronx, New York, from August 2005 until August 2007. The team included an advance practice nurse (APN) and social worker. A separate case—control study was conducted comparing the length of hospital stay for persons who died in the ICU during the final 6 months of the project, prior to and post-palliative care consultation for 22 patients at the hospital campus where the project team was located versus 24 patients at the other campus. Pharmaco-economic data were evaluated for 22 persons who died with and 43 who died without a palliative care consultation at the intervention campus ICU to evaluate whether the project intervention was associated with an increase in the use of pain medications or alterations in the use of potentially non-beneficial life-prolonging treatments in persons dying in the ICU. Data was abstracted from the medical record with a standardized chart abstraction instrument by an unblinded research assistant. Interviews were conducted with a sample of family members and ICU nurses rating the quality of end-of-life care in the ICU with the Quality of Dying and Death in the ICU instrument (ICUQODD), and a family focus group was also conducted. Forty percent of patients were Caucasian, 35% were African American or Afro-Caribbean, 22% Hispanic and 3% were Asian or other. Exploration of the patients’ and families’ needs identified significant spiritual needs in 62.4% of cases. Education on the death process was provided to 85% of families by the project team. Twenty-nine percent of patients were disconnected from mechanical ventilators following consultation with the Palliative Care Service (PCS), 15.9% of patients discontinued the use of inotropic support, 15.3% stopped artificial nutrition, 6.4% stopped dialysis and 2.5% discontinued artificial hydration. Recommendations on pain management were made for 51% of the project’s patients and symptom management for 52% of patients. The project was associated with an increase in the rate of the formalization of advance directives. Thirty-three percent of the patients who received PCS consultations had ‘do not resuscitate’ orders in place prior to consultation and 83.4% had ‘do not resuscitate’ orders after the intervention. The project team referred 80 (51%) of the project patients to hospice and 55 (35%) patients were enrolled on hospice, primarily at the medical center. The mean time from admission to palliative care consultation at the project site was 2.8 days versus 15.5 days at the other campus (p = 0.0184). Median survival times from admission to the medical center were not significantly different when stratified by palliative care consultation status: 12 days for the control group (95% CI 8—18) and 13.5 days for the intervention group (95% CI 8—20). Median charges for the use of opioid medications were higher (p = 0.01) for the intervention group but lower for use of laboratory (p = 0.004) and radiology tests (p = 0.027). We conclude that the integration of palliative care experts into the operation of critical care units is of benefit to patients, families and critical care clinicians. Preliminary evidence suggest that such models may be associated with improved quality of life, higher rates of formalization of advance directives and utilization of hospices, as well as lower use of certain non-beneficial life-prolonging treatments for critically ill patients who are at the end of life.

Impact of the AIDS epidemic on morbidity and mortality among intravenous drug users in a New York City methadone maintenance program.

To examine the impact of the AIDS epidemic on morbidity and mortality in a defined population of intravenous drug users, we analyzed overall and cause-specific death rates, AIDS incidence, and acute medical hospitalizations among patients in a long-term methadone maintenance program in New York City for the years 1984 through 1987 (midyear population for each year 828 to 891; demographic characteristics did not differ). The number of deaths while in treatment increased from 11 (13.3/1000) in 1984 to 39 (44.2/1000) in 1987. Deaths from AIDS increased from 3.6/1000 to 14.7/1000, deaths due to bacterial pneumonia/sepsis from 3.6/1000 to 13.6/1000; deaths from cirrhosis, drug overdose, trauma, and other causes remained relatively stable. AIDS incidence rose from six cases/1000 in 1984 to 20.4.1000 in 1987. Hospitalizations for AIDS, pneumonia, tuberculosis, and endocarditis/sepsis increased from 84.9/1000 in 1986 to 144.8/1000 in 1987. These data suggest that the AIDS epidemic has had a profound effect on patterns of morbidity and mortality among intravenous drug users in this methadone program population. Drug treatment programs may be important sites for targeting clinical services for drug users with AIDS, although the increasing burden of AIDS-related disease will require expansion of existing funding and treatment resources.

High prevalence of insomnia in an outpatient population with HIV infection.

OBJECTIVE To determine the prevalence, characteristics, and clinical recognition of insomnia in HIV-seropositive outpatients. DESIGN Systematic sample of patients attending an outpatient HIV/AIDS clinic at an urban teaching hospital. METHODS Patients were recruited for standard interviews including Pittsburgh Sleep Quality Index (PSQI); Mini Mental State Exam (MMSE); Trail Making Test A and B, and Hospital Anxiety and Depression Scale. Abstraction of medical records was used to assess clinician identification of insomnia. RESULTS 115 of 127 (91%) patients who were approached agreed to be interviewed for the study. Of these, 79 (69%) were male, 36 (31%) female; 38 (33%) black, 64 (56%) white, 12 (10%) Hispanic, 1 (1%) native American; 45 (39%) were injection drug users. Eighty-four (73%) respondents were classified as having a sleep disturbance according to the PSQI. Patients with cognitive impairment had a higher prevalence of insomnia (100% versus 70%; p = .034). A trend was shown for drug-using patients to report a higher prevalence of insomnia than nondrug users (86% versus 69%; p < .07). Using multivariate analysis, cognitive impairment (odds ratio [OR] = 1.4) as defined by the neuropsychiatric test battery and depression (OR = 1.2) were the best predictors of insomnia (p < .05). Only 28 (33%) patients with insomnia had any documentation of sleep disturbance in their clinical medical records. CONCLUSIONS Insomnia is widespread and underdiagnosed in HIV-seropositive ambulatory patients. Insomnia is especially prevalent among those with cognitive impairment. These findings suggest the importance for clinicians to inquire specifically about sleep disorders in HIV-seropositive patients. Prompt diagnosis and treatment may improve the quality of life in patients living with HIV.

An epidemic in evolution: the need for new models of HIV care in the chronic disease era.

Since the beginning of the AIDS epidemic, models of HIV care have needed to be invented or modified as the needs of patients and communities evolved. Early in the epidemic, primary care and palliative care predominated; subsequently, the emergence of effective therapy for HIV infection led to further specialization and a focus on increasingly complex antiretroviral therapy as the cornerstone of effective HIV care. Over the past decade, factors including (1) an aging, long-surviving population; (2) multiple co-morbidities; (3) polypharmacy; and (4) the need for chronic disease management have led to a need for further evolution of HIV care models. Moreover, geographic diffusion; persistent disparities in timely HIV diagnosis, treatment access, and outcomes; and the aging of the HIV provider workforce also suggest the importance of reincorporating primary care providers into the spectrum of HIV care in the current era. Although some HIV-dedicated treatment centers offer comprehensive medical services, other models of HIV care potentially exist and should be developed and evaluated. In particular, primary care- and community-based collaborative practices—where HIV experts or specialists are incorporated into existing health centers—are one approach that combines the benefits of HIV-specific expertise and comprehensive primary care using an integrated, patient-centered approach.

Clinical predictors of Pneumocystis carinii pneumonia, bacterial pneumonia and tuberculosis in HIV-infected patients

Background:Clinicians are frequently faced with the differential diagnosis between Pneumocystis carinii pneumonia (PCP), bacterial pneumonia, and pulmonary tuberculosis in HIV-infected patients. Objectives:To identify features that could help differentiate these three pneumonia types at presentation by evaluating the clinical characteristics of the three diagnoses among patients at two urban teaching hospitals. Design:Retrospective chart review. Methods:Cases were HIV-infected patients with a verified hospital discharge diagnosis of PCP (n = 99), bacterial pneumonia (n = 94), or tuberculosis (n = 36). Admitting notes were reviewed in a standardized manner; univariate and multivariate analyses were used to determine clinical predictors of each diagnosis. Results:Combinations of variables with the highest sensitivity, specificity, and odds ratios (OR) were as follows: for PCP, exertional dyspnea plus interstitial infiltrate (sensitivity 58%, specificity 92%; OR, 16.3); for bacterial pneumonia, lobar infiltrate plus fever ≤ 7 days duration (sensitivity 48%, specificity 94%; OR, 14.6); and for tuberculosis, cough > 7 days plus night sweats (sensitivity 33%, specificity 86%; OR, 3.1). On regression analysis, independent predictors included interstitial infiltrate (OR, 10.2), exertional dyspnea (OR, 4.9), and oral thrush (OR, 2.9) for PCP; rhonchi on examination (OR, 12.4), a chart mention of ‘toxic’ appearance (OR, 9.1), fever ≤ 7 days (OR, 6.6), and lobar infiltrate (OR, 5.8) for bacterial pneumonia; and cavitary infiltrate (OR, 21.1), fever > 7 days (OR, 3.9), and weight loss (OR, 3.6) for tuberculosis. Conclusions:Simple clinical variables, all readily available at the time of hospital admission, can help to differentiate these common pneumonia syndromes in HIV-infected patients. These findings can help to inform clinical decision-making regarding choice of therapy, use of invasive diagnostic procedures, and need for respiratory isolation.