Abstract
Objective
The aim of this study was to examine Emergency Department (ED) utilization and clinical and sociodemographic correlates of ED use among HIV-infected patients.
Methods
During 2003, 951 patients participated in face-to-face interviews at 14 HIV clinics in the HIV Research Network. Respondents reported the number of ED visits in the preceding 6 months. Using logistic regression, we identified factors associated with visiting the ED in the last 6 months and admission to the hospital from the ED.
Results
Thirty-two per cent of respondents reported at least one ED visit in the last 6 months. In multivariate analysis, any ED use was associated with Medicaid insurance, high levels of pain (the third or fourth quartile), more than seven primary care visits in the last 6 months, current or former illicit drug use, social alcohol use and female gender.
Of those who used ED services, 39% reported at least one admission to the hospital. Patients with pain in the highest quartile reported increased admission rates from the ED as did those who made six or seven primary care visits, or more than seven primary care visits vs. three or fewer.
Conclusions
The likelihood of visiting the ED has not diminished since the advent of highly active antiretroviraltherapy (HAART). More ED visits are to treat illnesses not related to HIV or injuries than to treat direct sequelae of HIV infection. With the growing prevalence of people living with HIV infection,the numbers of HIV-infected patients visiting the ED may increase, and ED providers need to understand potential complications produced by HIV disease.
Keywords: Emergency Department, HIV, utilization, HIV Research Network
Introduction
HIV-infected patients are more intensive users of the healthcare system than the general population.(1,2) Studies early in the HIV epidemic demonstrated that this population had a higher than average rate of Emergency Department (ED) use compared to the general United States population. (3) HIV-infected patients who present to the ED also have high rates of admission to the hospital. (4) In contrast to ED utilization in the general population, socio-demographic characteristics and drug use contributed little to the probability of ED visits in a cohort of HIV-infected persons receiving care in 1991–1992; ED was primarily driven by disease severity.(5)
Inpatient utilization has declined and outpatient utilization increased with the advent of HAART, but rates of ED utilization have not been reported in the current era of highly active antiretroviral therapy (HAART). (6–10) Emergency Department care is expensive and may be potentially avoidable. Identifying factors associated with ED visits is an important step in improving health care delivery to HIV-infected patients and reducing health care costs. Since HIV-infected patients are now living longer and healthier lives, (11–14) we hypothesized that ED utilization and inpatient admissions would be more strongly associated with sociodemographic and substance use characteristics, compared to factors related to the clinical aspects of HIV disease.(15–19)
The objective of this study was to assess utilization rates, reasons for ED utilization, and patient characteristics associated with ED utilization in the HAART era among patients who have a primary source of HIV care. We evaluated the characteristics associated with one or more ED visits, including demographic factors, frequency of primary care visits, pain, CD4 count and HIV-1 RNA. We also examined factors associated with being admitted to the hospital from the ED.
Methods
Study Design
This study is a cross-sectional survey, based on in-person interviews with patients recruited from HIV clinics. Patients were not recruited directly from Emergency Departments.
Study Setting
The HIV Research Network (HIVRN) is a consortium of outpatient clinics that provide primary and subspecialty care to HIV-infected adult and pediatric patients. Clinics abstract specified data elements from patients’ medical records; abstracted data are assembled into a uniform database and submitted to a Data Coordinating Center (2,20). Patients are identified only by a coded ID number in the medical record database. Fourteen out of the 15 clinics that treated adult patients participated in conducting interviews with patients. Six are located in the Eastern United States, three in the Midwest, two in the South and three in the West. Seven clinics have academic affiliations; seven are community-based.
Subject Recruitment
Initially, Data Coordinating Center staff drew a random sample from each participating clinic using the coded IDs in the medical record database. The sampling frame consisted of active patients in 2002 at these sites. Sampled IDs were then sent to the clinics to be linked with personal identifiers by clinic staff. Due to confidentiality restrictions, each sampled patient had to be first approached by a clinic staff member to solicit participation in the interview. Clinic staff mailed letters of invitation to potential study patients at their last known address. We encountered a large proportion of incorrect addresses and a high rate of non-response to the mailed letter of invitation. Resources did not permit multiple follow-ups of sampled patients, nor could it be documented whether nonresponse was due to incorrect addresses or to implicit refusal. Of 5,363 letters of invitation sent, we successfully conducted interviews with 717 patients (13%). To increase the sample size, in all but three clinics patients were recruited while awaiting treatment in the HIV clinic. This yielded interviews with another 234 patients. Time constraints on clinic staff precluded keeping detailed records of numbers of refusals, either to the letter or to the in-person recruitment. A total of 951 patients were interviewed. The median sample size per clinic was 59 patients (range: 38 to 172 patients). The low response rate to the mailed invitation, and the non-random selection of patients as they waited in clinics, implies that this should be considered a convenience sample. However, gender, race/ethnicity, the reported means of HIV acquisition, first CD4 count in 2003, and proportion with undetectable HIV-1 RNA were similar in the interviewed sample and in the larger population of patients at these clinics. (Table 1) The near-zero values of Cramer’s V statistic indicate very little association between data source and each variable.
Table 1.
Comparison of Interviewed Patients (N=871) and Full HIVRN Medical Record Data (n= 13,029)
| Overall Dataset Adult Patients Enrolled in 13 HIV Research Network Sites |
Interview Dataset | Cramer’s V Statistic | |
|---|---|---|---|
| Male Sex | 70% | 67% | −0.016 |
| Race | |||
| White | 29% | 28% | |
| Black | 49% | 56% | 0.048 |
| Hispanic | 20% | 13% | |
| Other/Unknown | 2.0% | 3.0% | |
| Transmission Method* | |||
| IDU | 25% | 28% | |
| MSM | 36% | 34% | 0.023 |
| HET | 33% | 30% | |
| First CD4 cell count | |||
| >50 | 8% | 6% | |
| 50–199 | 18% | 16% | |
| 200–499 | 43% | 45% | 0.018 |
| ≥500 | 31% | 33% | |
| HIV-1 RNA | |||
| <400 | 56% | 51% | |
| ≥400 | 44% | 49% | 0.026 |
IDU-Injection drug use, MSM- Men who have sex with men, HET- heterosexuals. For comparability, analyses exclude interview data from one site that did not provide medical record data in 2003. Missing data also excluded for comparability. Cramer’s V statistic measures strength of association and ranges from −1 to +1, with 0 indicating no association.
Face-to-face interviews were conducted between December 1, 2002 and December 31, 2003 by professional interviewers trained and supervised by Battelle Corporation (Columbus, Ohio, USA). The interviews assessed a wide range of HIV-related topics. For comparability, interview questions were taken from the interview developed for the HIV Cost and Services Utilization Study (HCSUS).(1,2) All patients in this study were receiving primary outpatient care, defined by having at least one CD4 test and one outpatient visit during 2003.
Institutional Review Board approval/exemption of the project, including the interview, was obtained by the data coordinating center and each clinic. Additionally, written informed consent was obtained from each participant before the start of the interview. Participants were reimbursed $30 for the approximately one-hour interview. A Spanish language version of the interview was available.
Measurement of Variables
Interview Data
The interview assessed the frequency of ED utilization in the prior 6 months, the number of ED visits that led to admission to the hospital, and whether the patient went to the ED on their own or on the advice of a healthcare provider. Patients were asked the reason for the most recent ED visit, with response options of: an illness you thought related to HIV infection, an accident or injury, pregnancy-related care, an alcohol or drug-related condition, an illness that was not related to HIV infection. We also examined HIVRN medical record data to determine the one year ED utilization rate among all adult patients enrolled in these HIVRN sites.
From the patient interview we collected data on gender, age, racial/ethnic group, years of school completed, employment status, insurance coverage, and HIV risk factor. (Table 1) We categorized age as 18–49 and 50 or older, based on the CDC guidelines for defining “elderly” HIV-infected individuals. For employment, respondents could self-identify as “disabled”; this does not imply that their disability status had been officially adjudicated. Respondents reported the number of primary care visits in the six months prior to the interview, excluding ED visits and excluding primary care visits solely for mental or substance abuse treatment. The number of visits was categorized into quartiles. Alcohol use was ascertained, as in HCSUS (22), from questions asking (1) how many days in the past four weeks the respondent drank alcohol, (2) how many drinks the person consumed on a typical day when drinking, and (3) the number of days the person consumed more than five drinks. We defined hazardous drinking as greater than 14 drinks per week for men and greater than 7 drinks per week for women, according to National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidelines.(23) Binge drinking was defined as 5 or more drinks on at least 1 day in the past 4 weeks. We combined hazardous and binge drinkers into one category, with the reference group being non-drinkers. “Social” drinkers were those who consumed alcohol, but not to excess.
To assess illicit drug use, we asked about use of sedatives, sleeping pills, tranquilizers, amphetamines, analgesics, marijuana, cocaine, inhalants, LSD, and heroin. Current substance use was defined as using any illicit drug within 6 months of the interview. Former substance use was defined as using illicit drugs more than 6 months prior to the interview, but not within 6 months of the interview.
Pain was measured by combining responses to the two items comprising the pain subscale of the Medical Outcome Study Short Form 36. Scoring was based on the RAND modifications to the SF-36. A score of zero represented no pain while a score of 100 represented the most intense pain.(24) For analyses, we classified this variable into quartiles.
Medical Record Data
CD4 count and HIV-1 RNA were extracted from medical records, using the first value obtained in calendar year 2003. CD4 count was categorized as 0–49, 50–199, 200–499 and greater than 499 cells/mm3. HIV-1 RNA was categorized as ≤400 copies, > 400 copies per/mL or as missing. HIV risk factor was also obtained from medical records; the IDU category comprised patients with multiple risk factors including IDU.
Highly active antiretroviral therapy (HAART) was defined as use of: (1) three or more nucleosides; or (2) a protease inhibitor [PI], a non-nucleoside reverse transcriptase inhibitor [RTI], or a fusion inhibitor, in combination with at least two other antiretrovirals. Patients were considered to be on HAART if they received any of these combinations during the calendar year. This definition includes triple nucleoside regimens, which were still standard of care in 2003. We opted to be as inclusive as possible in our definition of HAART in order to maximize the sensitivity of the analysis; this definition is unlikely to exclude any preferred drug combinations.
To compare the 6-month utilization rate with national statistics on ED use, we confirmed that the annual rate was twice the 6-month rate. We used HIVRN medical record data for all adult patients to determine ED visit rates for the first 6 months of 2003, the second six months of 2003, and the full year. Because ED use at providers outside the HIVRN may not be recorded in medial records, the ED visit rate obtained from medical record data may understate the true rate. However, there is no reason to believe that any potential undercount would vary differentially over time, and thus the medical record data can speak to relative rates for different time periods.
Data Analysis
We used chi-square tests to examine the association between individual socio-demographic variables and any ED use. Logistic regression was performed to analyze factors associated with having at least one visit to the ED, and with being admitted to the hospital from the ED. The multivariate model included variables presumed a priori to influence ED utilization. The Andersen-Aday model of the determinants of healthcare utilization provided the basis for our a priori assumptions. The model considers three sets of variables; predisposing characteristics, such as demographics, enabling factors, such as health insurance, and need factors, such as severity of current disease.(25,26) Multivariate analyses of any ED visit were conducted on 913 persons having complete data for all variables. The analyses of factors associated with hospital admission were conducted only among those who visited the ED and had no missing data. (N=280) Analyses were conducted using STATA 9.0, (College Station, Texas, USA). In all regressions, adjustment was made for site of care, to account for variations in practice patterns and demographic differences across clinics. This was done by adding an indicator variable for each clinic (except one reference clinic) to each model. All models were checked using likelihood ratio tests and the Hosmer-Lemeshow goodness of fit test.(27) For variables with multiple categories, we report as a “group test,” the Wald test for joint significance of all levels of the variable.
Results
The majority of the participants were male (68%) and of minority ethnicity (52% Black and 14% Hispanic) (Table 2). The median age was 45 years (range 20–85). HIV risk factors included men who have sex with men (MSM) (34%), heterosexual transmission (30%) and injection drug use (IDU) (27%). The majority (69%) were on HAART. As of the first test available for the patient in 2003, the median CD4 count was 376 cells/mm3 (range 0–2040 cells/mm3) and the median HIV-1 RNA was 461 copies/ml (range 0–750,000 copies/ml), with 37% being undetectable.
Table 2.
Demographic and Clinical Characteristics
| Study Population N=951 (%) |
Patients with any ED visits N=308 (%) |
Patients admitted from ED N=121 (%) |
|
|---|---|---|---|
| Age (years) | |||
| 18–49 | 677 (71) | 222 (72) | 87 (72) |
| >50 | 267 (28) | 83 (27) | 33 (27) |
| Missing | 7 (1) | 3 (1) | 1 (1) |
| Race | |||
| White | 294 (31) | 77 (25) | 25 (21) |
| Black | 491 (52) | 177 (57) | 76 (62) |
| Hispanic | 130 (14) | 39 (13) | 14 (12) |
| Other/missing | 36 (4) | 15 (5) | 6 (5) |
| Sex | |||
| Male | 648 (68) | 176 (57) | 69 (57) |
| Female | 303 (32) | 132 (43) | 52 (43) |
| HIV Risk Factor* | |||
| IDU | 256 (27) | 110 (36) | 44 (36) |
| MSM | 319 (34) | 73 (24) | 25 (21) |
| HET | 282 (30) | 96 (31) | 35 (29) |
| Other/Missing | 94 (10) | 29 (9) | 17 (14) |
| First CD4 Count of 2003 (cells/mm3) | |||
| <50 | 45 (5) | 18 (6) | 11 (9) |
| 50–199 | 114 (12) | 35 (11) | 18 (15) |
| 200–499 | 319 (34) | 99 (32) | 36 (30) |
| >499 | 234 (25) | 70 (23) | 21 (17) |
| Missing | 239 (25) | 86 (28) | 35 (29) |
| HIV-1 RNA(copies/ml) | |||
| <400 | 349 (37) | 123 (34) | 52 (42) |
| ≥ 400 | 360 (38) | 100 (39) | 36 (36) |
| Missing | 242 (25) | 85 (35) | 33 (39) |
| HAART | |||
| No | 292 (31) | 100 (32) | 42 (35) |
| Yes | 659 (69) | 208 (68) | 79 (65) |
| Bodily Pain | |||
| 1st Quartile | 206 (22) | 40 (13) | 10 (8) |
| 2nd Quartile | 241 (25) | 66 (22) | 21 (17) |
| 3rd Quartile | 208 (22) | 77 (25) | 28 (23) |
| 4th Quartile | 291 (31) | 122 (40) | 60 (50) |
| Missing | 5 (1) | 3 (1) | 2 (2) |
| Illicit drug use | |||
| Current | 344 (36) | 122 (40) | 48 (40) |
| Former | 336 (35) | 121 (40) | 48 (40) |
| Never | 267 (28) | 62 (22) | 24 (20) |
| Missing | 4 (0) | 3 (1) | 1 (1) |
| Alcohol use | |||
| Hazardous binge drinking | 101 (10) | 20 (7) | 6 (5) |
| Social Alcohol Usage | 275 (29) | 81 (27) | 25 (21) |
| No Alcohol | 573 (60) | 204 (67) | 89 (74) |
| Missing | 2 (1) | 1 (0) | 1 (1) |
| Education | |||
| < High school degree | 244 (26) | 92 (30) | 39 (32) |
| High school grad/junior college | 569 (61) | 186 (61) | 72 (59) |
| College/ post-college | 127 (14) | 27(9) | 9 (7) |
| Missing | 11 (1) | 3 (1) | 1 (1) |
| Employment | |||
| Working | 225 (24) | 48 (16) | 10 (8) |
| Retired | 151 (16) | 54 (17) | 25 (21) |
| Disabled | 475 (50) | 178 (58) | 77 (64) |
| Unemployed | 91 (10) | 26 (9) | 9 (7) |
| Missing | 9 (1) | 2 (1) | 0 (0) |
| Insurance | |||
| Private | 215 (23) | 51 (17) | 17 (14) |
| Medicaid | 357 (38) | 147 (48) | 58 (48) |
| Medicare | 92 (10) | 28 (9) | 15 (12) |
| Medicare/Medicaid | 181 (19) | 61 (20) | 25 (21) |
| None | 94 (10) | 16 (5) | 5 (4) |
| Missing | 12 (1) | 5 (2) | 1 (1) |
| Primary Care Visits | |||
| ≤3 | 375 (40) | 95 (31) | 24 (20) |
| 4, 5 | 117 (12) | 33 (11) | 9 (7) |
| 6, 7 | 227 (24) | 78 (26) | 39 (32) |
| ≥8 | 232 (24) | 99 (32) | 47 (40) |
| Number of ED Visits | |||
| 0 | 642 (68) | ---- | |
| 1 | 162 (17) | 52 (43) | |
| 2 | 74 (8) | 29 (24) | |
| 3 or more | 65 (7) | 40 (33) | |
| Missing | 8 (1) | ---- | |
IDU-Injection drug use, MSM- Men who have sex with men, , HET- heterosexual transmission.
Note: Percentages may not sum to 100 due to rounding.
ED Visits
The 6-month ED visit rate was 62.8 per 100 persons. Nearly one third (32%) reported at least one ED visit in the six months preceding the interview. Of those visiting the ED, the median was 1 visit per person (range 1–12). Of those with an ED visit, forty six percent made more than one visit in six months.
Per patient report, reasons for the most recent ED visit were to treat (1) an HIV-related illness (30%), (2) a non-HIV illness (45%), (3) an accident (14%), (4) drug or alcohol-related reasons (3%), and (5) pregnancy (0.3%), with 8% missing. For the most recent ED visit, over three quarters (77%) were self-referrals, and only 22% of visits were due to provider referral. For the most recent ED visit, twenty-six percent of those seeking emergency care for HIV-related illness were referred to the ED by the provider, and 22% of those seeking care for non-HIV-related illness were referred by the provider, a non-significant difference.
Table 3 presents results of a logistic regression analysis of factors associated with any ED visit, conducted on 913 patients with complete data. High levels of pain (third or fourth quartile), having more than seven primary care visits in the last six months, current or former illicit drug use, Medicaid insurance, and female gender remained associated with ED utilization when other variables were controlled. Clinical variables -- such as CD4 count, HIV-1 RNA, or HAART usage – were not significantly associated with any ED utilization.
Table 3.
Factors Associated with making any ED Visit within the past 6 months
| Bivariate OR (95% CI) Associated with an ED Visit |
Multivariate OR (95% CI) Associated with any ED Visit |
|
|---|---|---|
| Age (years) | ||
| 18–49 | 1.0 (Ref) | 1.0 (Ref) |
| ≥50 | 0.94 (0.69–1.28) | 0.95 (0.66–1.37) |
| Race | ||
| White | 1.0 (Ref) | 1.0 (Ref) |
| Black | 1.45 (0.98–2.13) | 1.27 (0.82–1.98) |
| Hispanic | 1.00 (0.60–1.67) | 0.94 (0.53–1.65) |
| Other | 1.72 (0.79–3.73) | 1.38 (0.59–3.23) |
| Group Test (P-Value) | 5.63 (0.13) | 2.21 (0.53) |
| Sex | ||
| Male | 1.0 (Ref) | 1.0 (Ref) |
| Female | 1.98 (1.46–2.69)* | 1.95 (1.33–2.86)* |
| HIV Risk Factor | ||
| IDU | 1.0 (Ref) | 1.0 (Ref) |
| MSM | 0.45 (0.31–0.66)* | 0.82 (0.52–1.29) |
| HET | 0.72 (0.50–1.04) | 0.67 (0.44–1.02) |
| Other/missing | 0.54 (0.31–0.91)* | 0.51 (0.28–0.92)* 6.40 (0.09) |
| CD4 in 2003 (cells/mm3) | ||
| <50 | 1.72 (0.87–3.44) | 1.41 (0.64–3.10) |
| 50–199 | 1.06 (0.64–1.74) | 0.92 (0.52–1.60) |
| 200–499 | 1.07 (0.74–1.57) | 0.99 (0.65–1.52) |
| >499 | 1.0 (Ref) | 1.0(Ref) |
| Missing | 1.21 (0.74–1.98) | 2.65 (0.58–12.11) |
| Group Test (P-Value) | 2.66 (0.61) | 2.64 (0.62) |
| HIV-1 RNA (copies/ml) | ||
| <400 | 1.0 (Ref) | 1.0 (Ref) |
| ≥400 | 0.85 (0.61 −1.17) | 1.05 (0.71 −1.55) |
| Missing | 0.97 (0.61 – 1.54) | 0.44 (0.10– 1.98) |
| Group Test (P-Value) | 1.05 (0.59) | 1.27 (0.53) |
| Receipt of HAART | ||
| No | 1.0 (Ref) | 1.0 (Ref) |
| Yes | 0.87 (0.64–1.17) | 0.99 (0.70–1.39) |
| Bodily Pain | ||
| 1st Quartile | 1.0 (Ref) | 1.0 (Ref) |
| 2nd Quartile | 1.56 (0.99–2.50) | 1.36 (0.83–2.23) |
| 3rd Quartile | 2.30 (1.45–3.63)* | 2.10 (1.28–3.44)* |
| 4th Quartile | 3.04 (1.98–4.68)* | 2.36 (1.47–3.81)* |
| Group Test (P-Value) | 29.36 (0.00) | 15.79 (0.001) |
| Illicit drug use | ||
| Current | 1.90 (1.30–2.77)* | 1.85 (1.20–2.86)* |
| Former | 1.82 (1.25–2.66)* | 1.59 (1.04–2.41)* |
| Never | 1.0 (Ref) | 1.0 (Ref) |
| Group Test (P-Value) | 12.85 (0.00) | 8.12 (0.02) |
| Alcohol use | ||
| Hazardous/Binge drinking | 1.0 (Ref) | 1.0 (Ref) |
| Social EtOH | 1.61 (0.93–2.81) | 1.76 (0.98–3.14) |
| No EtOH | 1.95 (1.16–3.28)* | 1.94 (1.06–3.57)* |
| Group Test (P-Value) | 6.81 (0.03) | 4.72 (0.09) |
| Education | ||
| < High School | 1.0 (Ref) | 1.0 (Ref) |
| H.S. or Junior College | 0.86 (0.62 −1.19) | 1.07 (0.74–1.54) |
| College or More | 0.53 (0.31 −0.89) | 0.85 (0.46–1.56) |
| Group Test (P-Value) | 6.13 (0.05) | 0.80 (0.67) |
| Employment | ||
| Working | 1.0 (Ref) | 1.0 (Ref) |
| Retired | 1.80 (1.10–2.96)* | 1.24 (0.71–2.19) |
| Disabled | 2.20 (1.50–3.22)* | 1.27 (0.80–2.04) |
| Unemployed | 1.31 (0.73–2.34) | 0.92 (0.48–1.77) |
| Group Test (P-Value) | 18.03 (0.00) | 1.87 (0.60) |
| Insurance | ||
| Private | 1.0 (Ref) | 1.0 (Ref) |
| Medicaid | 2.89 (1.78–4.70)* | 2.09 (1.20–3.65)* |
| Medicare | 1.15 (0.64–2.09) | 0.97 (0.51–1.87) |
| Medicaid/Medicare | 1.97 (1.18–3.29)* | 1.44 (0.81–2.57) |
| None | 0.76 (0.39–1.49) | 0.86 (0.41–1.80) |
| Group Test (P-Value) | 29.54 (0.00) | 11.30 (0.02) |
| Primary Care Visits past 6 months |
||
| ≤3 | 1.0 (Ref) | 1.0 (Ref) |
| 4, 5 | 1.09 (0.67–1.77) | 0.95 (0.56–1.61) |
| 6, 7 | 1.65 (1.11–2.45)* | 1.38 (0.90–2.13) |
| ≥8 | 2.63 (1.79–3.87)* | 2.13 (1.38–3.27)* |
| Group Test (P-Value) | 26.44 (0.00) | 14.28 (0.002) |
Note: Group test is a chi-square test of the joint significance of all category indicators for a variable. Analyses included indicators for each HIVRN site (not shown).
Hospital Admission from the ED
Thirty nine percent of patients who visited the ED (n=121) were subsequently admitted to the hospital from the ED on at least one occasion. The probability of having an admission from the ED was associated with the number of ED visits, rising from 32% of those with one ED visit, to 41% of those with 2 ED visits, to 67% of those with 3 or more visits (results not shown).
Table 4 reports results of a multivariate logistic regression of any inpatient admission from the ED (n= 280). The odds of admission to the hospital from the ED were greater for patients who made six or more primary care visits, versus 3 or fewer. Patients with CD4 counts less than 200 cells/mm3 were more likely to be admitted than those with CD4 counts above 500 cells/mm3. Patients reporting the highest level of pain also reported relatively high odds of admission from the ED, although the set of variables representing pain quartiles was not jointly significant. Patients who were retired had higher odds of being admitted from the ED than patients who were employed, but the overall effect of employment status on inpatient admissions was not significant.
Table 4.
Factors Associated with Admission to the Hospital from the ED (n=280)
| Variable Name | Bivariate Odds Ratio 95% (CI) |
Multivariate Odds Ratio 95% (CI) |
|---|---|---|
| Age | ||
| 18–49 | 1.0 (Ref) | 1.0 (Ref) |
| ≥50 | 0.95 (0.55–1.65) | 0.93 (0.45–1.91) |
| Race | ||
| White | 1.0 (Ref) | 1.0 (Ref) |
| Black | 1.71 (0.84–3.48) | 2.64 (1.00–6.95)* |
| Hispanic | 1.14 (0.44–2.98) | 1.14 (0.34–3.78) |
| Other | 1.26 (0.35–4.57) | 1.48 (0.28–7.77) |
| Group Test (P-Value) | 2.67 (0.44) | 5.28 (0.15) |
| Sex | ||
| Male | 1.0 (Ref) | 1.0 (Ref) |
| Female | 0.98 (0.59–1.62) | 1.10 (0.52–2.32) |
| HIV Risk Factor | ||
| IDU | 1.0 (Ref) | 1.0 (Ref) |
| MSM | 0.78 (0.42–1.48) | 1.02 (0.41–2.54) |
| HET | 0.92 (0.51–1.66) | 0.54 (0.25–1.20) |
| Other/missing | 1.88 (0.79–4.51) | 2.10 (0.68–6.50) |
| Group test (P-Value) | 3.58 (0.31) | 5.46 (0.14) |
| CD4 in 2003 (cells/mm3) | ||
| <50 | 3.60 (1.17–11.1)* | 5.74(1.29–25.52)* |
| 50–199 | 2.13 (0.90–5.07) | 2.91 (1.00–8.46)* |
| 200–499 | 1.34 (0.67–2.66) | 1.46 (0.62–3.46) |
| >499 | 1.0 (Ref) | 1.0 (Ref) |
| Missing | 2.20 (0.95–5.09) | 10.58 (0.99–112.06)* |
| Group Test (P-Value) | 7.95 (0.09) | 9.78 (0.04) |
| HIV-1 RNA (copies/ml) | ||
| <400 | 1.0 (Ref) | 1.0 (Ref) |
| ≥400 | 0.69 (0.39–1.21) | 0.79 (0.36–1.71) |
| Missing | 0.81 (0.45–1.47) | 0.14 (0.01–1.41) |
| Group Test (P-Value) | 1.72 (0.42) | 2.91 (0.23) |
| Receipt of HAART | ||
| No | 1.0 (Ref) | 1.0 (Ref) |
| Yes | 0.88 (0.53–1.46) | 1.28 (0.64–2.53) |
| Bodily Pain | ||
| 1st Quartile | 1.0 (Ref) | 1.0 (Ref) |
| 2nd Quartile | 1.42 (0.57–3.52) | 1.34 (0.43–4.18) |
| 3rd Quartile | 2.13 (0.88–5.17) | 1.80 (0.62–5.23) |
| 4th Quartile | 3.07 (1.34–7.00)* | 2.45 (0.88–6.82) |
| Group Test (P-Value) | 17.04 (0.00) | 3.62 (0.31) |
| Illicit drug use | ||
| Current | 1.11 (0.56–2.20) | 0.80 (0.31 – 2.04) |
| Former | 1.00 (0.50–2.00) | 0.71 (0.28 – 1.75) |
| Never | 1.0 (Ref) | 1.0 (Ref) |
| Group Test (P-Value) | 0.18 (0.95) | 0.57 (0.75) |
| Alcohol use | ||
| Hazardous/Binge drinking | 1.0 (Ref) | 1.0 (Ref) |
| Social EtOH | 1.29 (0.43–3.83) | 3.15 (0.83–11.94) |
| No EtOH | 2.03 (0.73–5.65) | 1.86 (0.47–7.35) |
| Group Test (P-Value) | 3.58 (0.16) | 3.70 (0.16) |
| Education | ||
| < High School | 1.26 (0.48–3.31) | 1.06 (0.53–2.12) |
| H.S. or Junior College | 1.24 (0.51–3.05) | 0.77 (0.23–2.61) |
| College or More | 1.0 (Ref) | 1.0 (Ref) |
| Group Test (P-Value) | 0.24 (0.88) | 0.33 (0.85) |
| Employment | ||
| Working | 1.0 (Ref) | 1.0 (Ref) |
| Retired | 2.89 (1.13–7.34)* | 4.27 (1.34–13.60)* |
| Disabled | 2.80 (1.28–6.11)* | 2.33 (0.85–6.41) |
| Unemployed | 2.00 (0.66–6.05) | 2.20 (0.54–8.97) |
| Group Test (P-Value) | 7.16 (0.07) | 6.04 (0.11) |
| Primary Care Visits past 6 months |
||
| <3 | 1.0 (Ref) | 1.0 (Ref) |
| 4, 5 | 1.38 (0.54–3.55) | 1.17 (0.35 – 3.93) |
| 6, 7 | 4.21 (2.02–8.78)* | 3.82 (1.61 −9.08)* |
| >8 | 3.27 (1.64–6.49)* | 3.16 (1.39 – 7.19)* |
| Group Test (P-Value) | 17.88 (0.00) | 11.50 (0.009) |
| Insurance | ||
| Private | 1.0 (Ref) | 1.0 (Ref) |
| Medicaid | 1.53 (0.61–3.85) | 0.56 (0.17 −1.87) |
| Medicare | 2.52 (0.88–7.21) | 1.32 (0.36 −4.85) |
| Medicare/Medicaid | 1.63 (0.63–4.20) | 0.72 (0.21 – 2.42) |
| None | 1.11 (0.31–4.07) | 0.83 (0.14– 4.83) |
| Group Test (P-Value) | 3.33 (0.50) | 2.16 (0.71) |
Note: Group test is a chi-square test of the joint significance of all category indicators for a variable. Analyses included indicators for each HIVRN site (not shown).
Discussion
ED utilization was high in this multiclinic, multistate sample of HIV-infected patients. In this study, 32% visited the ED once or more within six months, and the 6-month ED attendance rate was 62.8 per 100 persons. Inspection of HIVRN medical record data showed that the 1-year visit rate was approximately twice the 6-month rate. Extrapolation of the 6-month interview data leads to an estimated one-year ED visit rate of 125.6 per 100 persons. Nationwide United States estimates for the general population in 2003 show an overall ED visit rate of 38.9 visits per 100 persons, suggesting higher utilization among HIV-infected patients than in the general population. (3)
Our finding that nearly one-third of patients used the ED within a six-month interval is consistent with previous data from the pre-HAART and early HAART eras. In the pre- HAART era, 23%– 43% of HIV-infected persons reported utilizing the ED at least once in a six month period.(5,28–30) Other pre-HAART and early HAART estimates of ED utilization varied from 16 to 71% of patients studied over periods of three months to two years.(31–33) This suggests that the benefits of HAART have not resulted in decreased use of ED services.
A substantial proportion of patients (62%) reported that their most recent ED visit was due to non-HIV-related illnesses, injuries, or substance abuse. Future studies will need to examine provider-reported reasons for visitation, which may be different from patient self-report. Comparison of reason-for-visit data from patients and primary care providers may help to identify potentially avoidable ED visits. Nevertheless, if reasons for the most recent visit are representative of reasons for all visits, HIV infection may be incidental in a significant proportion of ED visits by people with HIV disease.
It is noteworthy that clinical variables, such as CD4 cell counts, HIV-1 RNA suppression, and receipt of HAART were not significantly associated with any ED use. This is in contrast to data from earlier in the HIV epidemic, showing that patients with AIDS were more likely to use the ED than those with less advanced HIV disease.(5) This finding is consistent with our hypothesis that ED use in the current HAART era will be less strongly related to clinical aspects of the disease, compared with the pre-HAART era.
Patients with more visits to the primary care physician were had higher odds of visiting the ED. Mauskopf et al., who found that patients with fewer than four visits had half the odds of making an ED visit, noted this association early in the epidemic.(34) One interpretation is that those with the most primary care visits are the most likely to use the ED because they are among the sickest patients. If true, illness burden may be related to comorbid conditions, given the lack of association between ED use and HIV-specific clinical variables. (The association between number of primary care visits and reason for the most recent ED visit was not significant.)
All subjects in this study were engaged with a source of regular HIV care. A previous study suggested that lacking a dominant HIV provider may increase the odds of using the ED (34), and thus our results may underestimate ED use by HIV-infected patients lacking a regular source of care. However, it is notable that ED use was relatively common in this sample, despite the presence of established links with primary care providers.
As hypothesized, ED utilization was associated with the same sociodemographic and substance use characteristics that are associated with increased ED use in the general population(35,36). Women, those with Medicaid insurance, and those who described themselves as disabled were more likely to use the ED than their counterparts. Many studies have demonstrated increased health care utilization in HIV-infected women compared to men.(2,29,37) Our findings are consistent with those of the HCSUS, which showed that women had more use of the ED than men. (29) While other studies have shown no differences in ED utilization between HIV-infected men and women, they generally examined subgroups of HIV-infected persons, particularly the homeless (30,38) or drug users (30,39) or used data from early in the HIV epidemic. HCSUS showed higher odds of ED use among persons with public insurance, racial/ethnic minorities, persons with IDU HIV exposure, and those under 35 years of age, whereas the current study did not find significant effects for age, HIV risk factor, or minority status. (29)
Like Solomon and Palepu, we found that both current and former drug users had higher odds of using the ED than those who never used drugs.(30,33) This could be because current drug users may have medical complications of injection drug use, such as abscesses, osteomyeletitis, endocarditis, and overdoses requiring emergency evaluation. Former drug users may have increased need for emergency services due to long-term sequelae of former drug use such as complications of infectious hepatitis. Although Palacio et al. did not find that IDU was associated with ED use among Women’s Interagency Health Study (WIHS) participants, our definition of illicit drug use was more inclusive than IDU/non-IDU, as we included patients who were using any illicit drug, independent of injection status. (31)
Consistent with past literature, (5,40) we found that higher levels of pain were associated with increased likelihood of ED utilization. The effect of pain was notable, given that it is possible that some ED visits could have occurred prior to the period (past four weeks) captured in the pain questions. The pain questions may be reflecting chronic pain that persists over periods longer than four weeks.
Thirty-nine percent of ED users had at least one inpatient hospitalization following ED visitation. This is consistent with several other serious chronic diseases and demonstrates significant severity of illness among HIV infected patients. Therefore, utilization of the ED utilization may therefore be appropriate in many instances.
Limitations
Results of this study should be interpreted in light of several limitations. First, we were limited by self-reported measures of ED utilization in this analysis. It is possible that some respondents forgot to include some ED visits in the total, while others may have reported visits that occurred outside the six-month reference period. However, reliance on self-reported data greatly increases the feasibility of studies of ED use, as medical records from one provider may not capture all ED use by a patient.
A large proportion of patients had missing CD4 cell count and HIV-1 RNA data. For 80 patients, data were missing because one site left the HIVRN after interviews were conducted and no medical record data were available for 2003. For others, a match with medial record data could not be established. Although patients with missing clinical data were included in analyses, the rate of missing data is a limitation.
In addition, the convenience sample of interviewees may introduce bias into the estimates of ED use, as respondents and non-respondents may differ in service use. Patients who were approached in the waiting room to participate may have differed from those who responded to the mailed invitation. This may also introduce bias concerning the number of visits to the HIV clinic. We compared all patients enrolled in the HIVRN during 2003 to those who participated in the interview and found no differences in gender, race, or HIV risk factor; however, there may still be other differences between those patients who chose to participate in the study and the overall population of patients using HIVRN clinics. The high percentage of interviewees who were unemployed, disabled, or retired may also have led to the introduction of bias, as these patients had more potential free time to attend an interview. Finally, the HIVRN is not a national probability sample. Though its population is similar to that of a 1996 nationally representative sample of persons in care for HIV infection (1), we are cautious about generalizing our findings to the entire U.S. HIV-infected population.
Conclusions
In summary, HIV-infected individuals make frequent visits to the Emergency Department and are often admitted from there to the hospital. The proportion of patients making one or more ED visits has apparently not declined since the introduction of HAART. The increased prevalence of patients with HIV infection due to improved survival with HAART, the aging of the population and the development of comorbid disease in HIV infected patients suggest that overall numbers of persons with HIV infection using ED services may be increasing over time. Although some ED visits are due to injuries, the majority are due to significant HIV or non-HIV related illnesses and the presence of HIV infection may complicate care delivery. ED providers need to be aware of the side effects of treatments and the management of comorbidities in HIV-infected patients. If pain management and substance abuse complications are associated with increased likelihood of ED visits, additional services to provide patients with adequate outpatient pain management and substance abuse treatment may reduce ED utilization. Our results are important not only for HIV-infected patients and providers but also for those who pay for this care.
Acknowledgments
Funding Sources: Supported by the Agency for Healthcare Research and Quality (290-01-012) and the National Institutes of Drug Abuse (K23-DA00523) and Aging (R01 AG026250). Dr. Gebo also received support from the Johns Hopkins University Richard S. Ross Clinician Scientist Award.
APPENDIX
Participating Sites
Alameda County Medical Center, Oakland, California (Howard Edelstein, MD, Silver Sisneros, D.O.)
Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania (Richard Rutstein, M.D.)
Community Health Network, Rochester, New York (Steven Fine, M.D., Roberto Corales, D.O.)
Community Medical Alliance, Boston, Massachusetts (James Hellinger, M.D.)
Drexel University, Philadelphia, Pennsylvania (Peter Sklar, M.D., Sara Allen, C.R.N.P.)
Henry Ford Hospital Detroit, Michigan (John Jovanovich, M.D., Norman Markowitz, M.D.)
Johns Hopkins University, Baltimore, Maryland (Kelly Gebo, M.D., Richard Moore, M.D., George Siberry MD, Allison Agwu MD)
Montefiore Medical Group, Bronx, New York (Robert Beil, M.D.)
Montefiore Medical Center, Bronx, New York (Lawrence Hanau, M.D.)
Nemechek Health Renewal, Kansas City, Missouri (Patrick Nemechek, M.D.)
Oregon Health and Science University, Portland, Oregon (P. Todd Korthuis, M.D.)
Parkland Health and Hospital System, Dallas, Texas (Philip Keiser, M.D.)
St. Jude’s Children’s Hospital and University of Tennessee, Memphis, Tennessee (Patricia Flynn, M.D., Aditya Gaur, M.D.)
St. Luke’s Roosevelt Hospital Center, New York, New York (Victoria Sharp, M.D.)
Tampa General Health Care, Tampa, Florida (Jeffrey Nadler, M.D., Chararut Somboonwit, M.D.)
University of California, San Diego, La Jolla, California (Stephen Spector, M.D.)
University of California, San Diego, California (W. Christopher Mathews, M.D.)
Wayne State University, Detroit, Michigan (Lawrence Crane, M.D., Jonathan Cohn, M.D.)
Sponsoring Agencies
Agency for Healthcare Research and Quality, Rockville, Maryland (Fred Hellinger, Ph.D., John Fleishman, Ph.D., Irene Fraser, Ph.D.)
Health Resources and Services Administration, Rockville, Maryland (Richard Conviser, Ph.D., Alice Kroliczak, Ph.D., Robert Mills, Ph.D.)
Substance Abuse and Mental Health Services Administration, Rockville, Maryland (Joan Dilonardo, Ph.D., Laura House, Ph.D., Pat Roth)
Data Coordinating Center
Johns Hopkins University (Richard Moore, M.D., Jeanne Keruly, C.R.N.P., Kelly Gebo, M.D., Perrin Lawrence, M.P.H., Alanna Zhao, M.S., Michelande Ridore, B.S.)
Footnotes
Prior Presentation: This paper was presented in poster form at the Conference on Retroviruses and Opportunistic Infection, Denver, Colorado February, 2006
Disclaimer: The views expressed in this paper are those of the authors. No official endorsement by DHHS, the National Institutes of Health, or the Agency for Healthcare Research and Quality is intended or should be inferred.
References
- 1.Bozzette SA, Berry SH, Duan N, Frankel MR, Leibowitz AA, Lefkowitz D, et al. The care of HIV-infected adults in the United States. HIV Cost and Services Utilization Study Consortium. N.Engl.J.Med. 1998 Dec 24;339(26):1897–1904. doi: 10.1056/NEJM199812243392606. [DOI] [PubMed] [Google Scholar]
- 2.Fleishman JA, Gebo KA, Reilly ED, Conviser R, Christopher Mathews W, Todd Korthuis P, et al. Hospital and outpatient health services utilization among HIV-infected adults in care 2000–2002. Med.Care. 2005 Sep;43(9 Suppl):III40–III52. doi: 10.1097/01.mlr.0000175621.65005.c6. [DOI] [PubMed] [Google Scholar]
- 3.McCaig L, MPH, Burt C., EdD [Accessed 8/16, 2007];National Hospital Ambulatory Medical Care Survey:2003 Emergency Department Summary. 2005 Available at: cdc.gov/nchs/data/ad/ad358.pdf. [PubMed]
- 4.Derlet RW, Richards JR. Overcrowding in the nation's emergency departments: complex causes and disturbing effects. Ann.Emerg.Med. 2000 Jan;35(1):63–68. doi: 10.1016/s0196-0644(00)70105-3. [DOI] [PubMed] [Google Scholar]
- 5.Pezzin LE, Fleishman JA. Is outpatient care associated with lower use of inpatient and emergency care? An analysis of persons with HIV disease. Acad.Emerg.Med. 2003 Nov;10(11):1228–1238. doi: 10.1111/j.1553-2712.2003.tb00607.x. [DOI] [PubMed] [Google Scholar]
- 6.Fleishman JA, Hellinger FH. Recent trends in HIV-related inpatient admissions 1996–2000: a 7-state study. J.Acquir.Immune Defic.Syndr. 2003 Sep 1;34(1):102–110. doi: 10.1097/00126334-200309010-00015. [DOI] [PubMed] [Google Scholar]
- 7.Palella FJ, Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N.Engl.J.Med. 1998 Mar 26;338(13):853–860. doi: 10.1056/NEJM199803263381301. [DOI] [PubMed] [Google Scholar]
- 8.Gebo KA, Diener-West M, Moore RD. Hospitalization rates differ by hepatitis C satus in an urban HIV cohort. J.Acquir.Immune Defic.Syndr. 2003 Oct 1;34(2):165–173. doi: 10.1097/00126334-200310010-00006. [DOI] [PubMed] [Google Scholar]
- 9.Torres RA, Barr M. Impact of combination therapy for HIV infection on inpatient census. N.Engl.J.Med. 1997 May 22;336(21):1531–1532. doi: 10.1056/nejm199705223362118. [DOI] [PubMed] [Google Scholar]
- 10.Mouton Y, Alfandari S, Valette M, Cartier F, Dellamonica P, Humbert G, et al. Impact of protease inhibitors on AIDS-defining events and hospitalizations in 10 French AIDS reference centres. Federation National des Centres de Lutte contre le SIDA. AIDS. 1997 Oct;11(12):F101–F105. doi: 10.1097/00002030-199712000-00003. [DOI] [PubMed] [Google Scholar]
- 11.Sterne JA, Hernan MA, Ledergerber B, Tilling K, Weber R, Sendi P, et al. Long-term effectiveness of potent antiretroviral therapy in preventing AIDS and death: a prospective cohort study. Lancet. 2005 Jul;366(9483):378–384. doi: 10.1016/S0140-6736(05)67022-5. 30-Aug 5. [DOI] [PubMed] [Google Scholar]
- 12.Hammer SM, Squires KE, Hughes MD, Grimes JM, Demeter LM, Currier JS, et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team. N.Engl.J.Med. 1997 Sep 11;337(11):725–733. doi: 10.1056/NEJM199709113371101. [DOI] [PubMed] [Google Scholar]
- 13.Egger M, Hirschel B, Francioli P, Sudre P, Wirz M, Flepp M, et al. Impact of new antiretroviral combination therapies in HIV infected patients in Switzerland: prospective multicentre study. Swiss HIV Cohort Study. BMJ. 1997 Nov 8;315(7117):1194–1199. doi: 10.1136/bmj.315.7117.1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jacobson LP, Li R, Phair J, Margolick JB, Rinaldo CR, Detels R, et al. Evaluation of the effectiveness of highly active antiretroviral therapy in persons with human immunodeficiency virus using biomarker-based equivalence of disease progression. Am.J.Epidemiol. 2002 Apr 15;155(8):760–770. doi: 10.1093/aje/155.8.760. [DOI] [PubMed] [Google Scholar]
- 15.French MT, McGeary KA, Chitwood DD, McCoy CB. Chronic illicit drug use, health services utilization and the cost of medical care. Soc.Sci.Med. 2000 Jun;50(12):1703–1713. doi: 10.1016/s0277-9536(99)00411-6. [DOI] [PubMed] [Google Scholar]
- 16.Alexandre PK, Roebuck MC, French MT, Chitwood DD, McCoy CB. Problem drinking, health services utilization, and the cost of medical care. Recent Dev.Alcohol. 2001;15:285–298. doi: 10.1007/978-0-306-47193-3_16. [DOI] [PubMed] [Google Scholar]
- 17.Weber EJ, Showstack JA, Hunt KA, Colby DC, Grimes B, Bacchetti P, et al. Are the Uninsured Responsible for the Increase in Emergency Department Visits in the United States? Ann.Emerg.Med. 2008 Apr 11; doi: 10.1016/j.annemergmed.2008.01.327. [DOI] [PubMed] [Google Scholar]
- 18.Hunt KA, Weber EJ, Showstack JA, Colby DC, Callaham ML. Characteristics of frequent users of emergency departments. Ann.Emerg.Med. 2006 Jul;48(1):1–8. doi: 10.1016/j.annemergmed.2005.12.030. [DOI] [PubMed] [Google Scholar]
- 19.Sun BC, Burstin HR, Brennan TA. Predictors and outcomes of frequent emergency department users. Acad.Emerg.Med. 2003 Apr;10(4):320–328. doi: 10.1111/j.1553-2712.2003.tb01344.x. [DOI] [PubMed] [Google Scholar]
- 20.Gebo KA, Fleishman JA, Reilly ED, Moore RD HIV Research Network. High rates of primary Mycobacterium avium complex and Pneumocystis jiroveci prophylaxis in the United States. Med.Care. 2005 Sep;43(9 Suppl):III23–III30. doi: 10.1097/01.mlr.0000175631.34438.1e. [DOI] [PubMed] [Google Scholar]
- 21.1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Recomm Rep. 1992 Dec 18;41(RR-17):1–19. [PubMed] [Google Scholar]
- 22.Galvan FH, Bing EG, Fleishman JA, London AS, Caetano R, Burnam MA, et al. The prevalence of alcohol consumption and heavy drinking among people with HIV in the United States: results from the HIV Cost and Services Utilization Study. J.Stud.Alcohol. 2002 Mar;63(2):179–186. doi: 10.15288/jsa.2002.63.179. [DOI] [PubMed] [Google Scholar]
- 23.National Institute on Alcohol Abuse and Alcoholism. Helping Patients Who Drink Too Much:A Clinicians Guide. 2005 [Google Scholar]
- 24.Scoring Instructions for the 36-Item Short Form Survey (SF-36) [Accessed 3/20, 2008];2007 Available at: http://rand.org/health/surveys_tools/mos/mos_core_36item_scoring.html.
- 25.Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J.Health Soc.Behav. 1995 Mar;36(1):1–10. [PubMed] [Google Scholar]
- 26.Andersen RM, McCutcheon A, Aday LA, Chiu GY, Bell R. Exploring dimensions of access to medical care. Health Serv.Res. 1983 Spring;18(1):49–74. [PMC free article] [PubMed] [Google Scholar]
- 27.Hosmer D, Lemeshow S. Applied Logistic Regression. 2nd ed. New York: Wiley & Sons; 2000. [Google Scholar]
- 28.Bailey JE, Van Brunt DL, Somes GW, Dorko CS, Wan JY. Impact of a statewide Medicaid managed care system on healthcare utilization and outcomes for people living with HIV. Am.J.Med.Sci. 2004 Dec;328(6):305–314. doi: 10.1016/s0002-9629(15)33939-2. [DOI] [PubMed] [Google Scholar]
- 29.Shapiro MF, Morton SC, McCaffrey DF, Senterfitt JW, Fleishman JA, Perlman JF, et al. Variations in the care of HIV-infected adults in the United States: results from the HIV Cost and Services Utilization Study. JAMA. 1999 Jun 23–30;281(24):2305–2315. doi: 10.1001/jama.281.24.2305. [DOI] [PubMed] [Google Scholar]
- 30.Palepu A, Horton NJ, Tibbetts N, Dukes K, Meli S, Samet JH. Substance abuse treatment and emergency department utilization among a cohort of HIV-infected persons with alcohol problems. J.Subst.Abuse Treat. 2003 Jul;25(1):37–42. doi: 10.1016/s0740-5472(03)00064-3. [DOI] [PubMed] [Google Scholar]
- 31.Palacio H, Li X, Wilson TE, Sacks H, Cohen MH, Richardson J, et al. Healthcare use by varied highly active antiretroviral therapy (HAART) strata: HAART use, discontinuation, and naivety. AIDS. 2004 Mar 5;18(4):621–630. doi: 10.1097/00002030-200403050-00006. [DOI] [PubMed] [Google Scholar]
- 32.Palacio H, Shiboski CH, Yelin EH, Hessol NA, Greenblatt RM. Access to and utilization of primary care services among HIV-infected women. J.Acquir.Immune Defic.Syndr. 1999 Aug 1;21(4):293–300. doi: 10.1097/00126334-199908010-00006. [DOI] [PubMed] [Google Scholar]
- 33.Solomon L, Stein M, Flynn C, Schuman P, Schoenbaum E, Moore J, et al. Health services use by urban women with or at risk for HIV-1 infection: the HIV Epidemiology Research Study (HERS) J.Acquir.Immune Defic.Syndr.Hum.Retrovirol. 1998 Mar 1;17(3):253–261. doi: 10.1097/00042560-199803010-00011. [DOI] [PubMed] [Google Scholar]
- 34.Mauskopf J, Turner BJ, Markson LE, Houchens RL, Fanning TR, McKee L. Patterns of ambulatory care for AIDS patients, and association with emergency room use. Health Serv.Res. 1994 Oct;29(4):489–510. [PMC free article] [PubMed] [Google Scholar]
- 35.Sun BC, Burstin HR, Brennan TA. Predictors and outcomes of frequent emergency department users. Acad.Emerg.Med. 2003 Apr;10(4):320–328. doi: 10.1111/j.1553-2712.2003.tb01344.x. [DOI] [PubMed] [Google Scholar]
- 36.Fryer GE, Jr, Green LA, Dovey SM, Yawn BP, Phillips RL, Lanier D. Variation in the ecology of medical care. Ann.Fam.Med. 2003 Jul–Aug;1(2):81–89. doi: 10.1370/afm.52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Gebo KA, Fleishman JA, Conviser R, Reilly ED, Korthuis PT, Moore RD, et al. Racial and gender disparities in receipt of highly active antiretroviral therapy persist in a multistate sample of HIV patients in 2001. J.Acquir.Immune Defic.Syndr. 2005 Jan 1;38(1):96–103. doi: 10.1097/00126334-200501010-00017. [DOI] [PubMed] [Google Scholar]
- 38.Masson CL, Sorensen JL, Phibbs CS, Okin RL. Predictors of medical service utilization among individuals with co-occurring HIV infection and substance abuse disorders. AIDS Care. 2004 Aug;16(6):744–755. doi: 10.1080/09540120412331269585. [DOI] [PubMed] [Google Scholar]
- 39.Knowlton AR, Hoover DR, Chung SE, Celentano DD, Vlahov D, Latkin CA. Access to medical care and service utilization among injection drug users with HIV/AIDS. Drug Alcohol Depend. 2001 Sep 1;64(1):55–62. doi: 10.1016/s0376-8716(00)00228-3. [DOI] [PubMed] [Google Scholar]
- 40.Fleishman JA, Hsia DC, Hellinger FJ. Correlates of medical service utilization among people with HIV infection. Health Serv.Res. 1994 Dec;29(5):527–548. [PMC free article] [PubMed] [Google Scholar]