International Journal for Quality in Health Care 16:41-50 (2004)
© International Society for Quality in Health Care and Oxford University Press 2004; all rights reserved
Symptom-based framework for assessing quality of HIV care
1 VA Greater Los Angeles Health System and the University of California, Los Angeles, CA,
2 RAND Health, Santa Monica, CA,
3 Division of General Medicine, University of Pennsylvania, Philadelphia, PA,
4 Health Services Research and Development Program, VA San Diego Healthcare System, San Diego, CA,
5 Division of Infectious Diseases, Department of Medicine,
6 UCSD School of Medicine, La Jolla, CA,
7 Division of General Internal Medicine and Health Services Research, School of Medicine, University of California, Los Angeles, CA, USA
Objective. To evaluate HIV quality of care using a symptom-based, patient-centered framework.
Methods. An expert panel developed 13 quality indicators for three common symptoms: cough with fever and/or shortness of breath; severe or persistent diarrhea; and significant weight loss. A nationally representative probability sample of HIV-infected adults was interviewed between 1996 and 1997. Participants were asked about the presence and severity of HIV symptoms during the preceding 6 months, and care received. Variation in adherence to the indicators was assessed by symptom type and patient characteristics.
Results. In all, 2864 (71%) patients completed interviews and 920 reported being at least moderately bothered with one of the three symptoms. Of these, 41, 74, and 65% of patients with a symptom of cough, weight loss, or diarrhea, respectively, reported receiving all indicated care. Performance was better for patients with more severe HIV, measured as a CD4 cell count <50 cells/microliter, compared with those with less severe HIV, measured as CD4 cell count >500 cells/microliter (43% versus 60%; P = 0.02). Uninsured patients had worse performance than Medicare patients (45% versus 62%; P = 0.04), but care did not differ by patient’s age, gender, ethnicity, HIV risk factor, providers’ HIV patient load, or region. Only CD4 count remained significantly associated with performance in the multivariate analyses.
Conclusions. Symptom-based quality indicators may provide a useful supplement to conventional measures. Patients with HIV reported substantial underuse of services for common, burdensome symptoms. Although adherence to quality indicators was better for patients with more advanced HIV disease, many still received suboptimal care. Vulnerable patient groups generally did not receive worse quality of care, suggesting that symptom-based measures of quality may measure domains that are distinct from those captured by conventional indicators.
Keywords: HIV, patient-centered care, quality indicators, quality of care, symptoms
Accepted for publication September 24, 2003.
Symptom assessment and care is an essential component of medical practice, yet until recently research on symptoms has received relatively little attention in leading medical journals [1–4]. Symptoms often prompt patients to seek medical attention and provide physicians with essential clues about change in health status and underlying pathology. More intense or bothersome symptoms, in particular, are generally expected to lead physicians to conduct a systematic evaluation and, if possible, to treat the most likely cause of the symptoms. Although physicians may vary in their approach to assessing a given symptom, there is a reasonable level of consensus regarding the minimal types of assessment and treatment for many common symptoms among patients with a known disease, such as HIV infection.
With the advent of more effective treatments, HIV has been transformed from a progressive and ultimately fatal disease into a reversible, chronic disease, with the possibility of long-term survival for many. Consequently, although preventing progression of HIV and minimizing secondary complications continue to be primary goals of care, controlling constitutional symptoms associated with HIV has emerged as a central goal, since such symptoms can significantly impact on health-related quality of life [4–11]. Such symptoms can also be an indicator of HIV progression with development of serious secondary complications, particularly among those at later stages of disease [1,4].
Despite increased recognition of the importance of addressing symptoms among HIV patients, few studies examine HIV-specific quality of care in terms of appropriate management of symptoms or have conceptualized symptoms that patients present with as the starting point to assess whether appropriate care was given [12,13]. Most quality of HIV care studies focus on physician-defined need for care, such as patients being prescribed appropriate antiretroviral therapy or opportunistic infection prophylaxis in the setting of depressed CD4 cell count, which indicates relatively advanced HIV disease. This work has shown that vulnerable populations (e.g. African Americans) who tend to receive lower quality of care in general also receive poorer HIV care [14,15]. In this study, we use nationally representative data from the HIV Cost and Services Utilization Study (HCSUS) to evaluate the quality of care provided for three types of common, bothersome symptom complexes among HIV patients: complicated cough, weight loss, and diarrhea. We also examine whether vulnerable populations receive poorer care for these symptoms.
 |
Methods |
|---|
 Top Methods ResultsDiscussionReferences |
|---|
Study population
The HCSUS reference population included persons aged 18 years or older with known HIV infection who made at least one visit, in the context of regular or ongoing care, to a non-military, non-prison medical provider (other than an emergency department) from January 1996 to February 1997. Full details of the HCSUS sampling design are presented elsewhere [16]. In brief, the HCSUS used a three-stage sampling design, in which geographical areas, medical providers, and patients were sampled. In the first stage, we sampled 28 metropolitan areas and 25 clusters of rural counties within the USA. In the second stage, we sampled 148 urban and 51 rural providers. In the third stage, we sampled 4042 patients from identified lists of all eligible patients who visited participating providers during January and February 1996. Of the 4042 patients were enrolled in the study, 2864 (71%) completed full interviews. Baseline interviews began in January 1996 and ended 15 months later. Ninety-one per cent of the 2864 interviews were conducted in person, and the remainder by telephone. Although there were several more recent follow-up interviews of the HCSUS sample, these did not include a sufficiently detailed set of questions about symptoms to use for the analyses reported here.
Symptom measures
All participants were asked if they had any of the following symptoms within the past 6 months: persistent cough, difficulty breathing, or difficulty catching your breath that lasted for more than 1 week; fever, shaking chills, or sweats; diarrhea, or loose or watery stools; or substantial weight loss (
10% of usual body weight). These symptoms were selected based on previous research [17] and expert judgment from clinicians and HIV community representatives. Interviewees experiencing a symptom were asked to report their perceived level of bother for each symptom based on a 5-point Likert-type scale: 0 – ‘not at all’, 1 – ‘very little’, 2 – ‘moderately’, 3 – ‘quite a bit’, and 4– ‘extremely’.
In this paper, we focus on care for patients presenting with three types of symptoms. The first type, referred to as complicated cough, includes a persistent cough associated with shortness of breath (SOB) and/or fever, shaking chills, and sweats. The other two symptom types are diarrhea and weight loss. We selected these symptoms because they are indicative of potentially dangerous underlying conditions such as Pneumocystis carinii pneumonia (cough) [16] or intestinal infections (diarrhea) [17], or other disseminated opportunistic infections or tumors that adversely affect quality of life and survival (weight loss) [18,19].
Participants who reported that they were at least ‘moderately’ bothered by cough, weight loss, or diarrhea within the past 6months and had obtained care were selected for analysis. If patients experienced more than one of these three symptoms, they were asked to choose which one was the most bothersome. We studied care for the most bothersome symptom, because it is likely to reflect the individual’s greatest need for care.
Development of quality indicators
A panel of 10 HIV clinician experts, including most of the authors of this paper, was convened during the development of the HCSUS baseline questionnaire to identify specific processes of care that a clinician would be expected to perform for HIV patients presenting with particular symptoms. In choosing indicators, panel members assumed that patients with more symptomatic HIV had more advanced disease. The emphasis was not on identifying a comprehensive list of indicators for a given symptom, but rather on identifying basic processes of care that could be reliably reported by patients. Proposed indicators were circulated, revised after discussion, and retained only if consensus was achieved. Panelists identified a total of 13 separate indicators of quality (Table 1) for the three symptom types, including five for complicated cough, five for weight loss, and three for diarrhea. Panelists considered including highly active antiretroviral therapy (HAART) or viral load in the indicators of quality of care, but decided that these technologies were too rapidly diffusing in the course of the study to be included. CD4 count was included as a measure of severity in some indicators as well as a covariate in our analyses.
|
For example, all patients presenting with complicated cough would be expected to have their lungs examined. Patients with cough and a CD4 count <200 cells/microliter would also be expected to have a chest X-ray and an induced sputum or bronchoscopy. Similarly, all patients with diarrhea would be expected to have stool studies, and those with persistent diarrhea unresponsive to treatment would be expected to have an endoscopy.
Statistical analyses
Patient reports about receiving specific processes of care were used to calculate an overall performance rate for each type of symptom. We first created dichotomous variables that indicated whether or not an expected process of care was performed. We then calculated the percentage of patients with a given symptom who received all indicated types of care for that symptom. In addition, we calculated an overall performance rate defined as the percentage of patients who received all indicated care for any of the three types of symptoms combined. For descriptive purposes, we also categorized the indicators into five distinct types and calculated the percentage of patients receiving all indicated care for each of the following types of indicators: history, physical exam, non-invasive procedure, invasive procedure, and treatment.
We compared the sociodemographic and clinical characteristics of patients who experienced any of the three types of bothersome symptoms with patients who did not experience these symptoms using 
2-tests for categorical variables and a Wilcoxon rank sum test for ordinal variables. For each type of symptom, we also compared the percentage of patients in different sociodemographic and clinical subgroups receiving all indicated care. We used multivariate logistic regression to assess whether specific sociodemographic characteristics or clinical factors were associated with indicated types of care when other covariates were taken into account. Covariates included CD4 cell count, age, gender, race/ethnicity, education, insurance type, exposure history, region, and the HIV patient load of the site at which patients received care.
All analyses were weighted to represent the national population under care for HIV disease. Weights were based on: (i) different probabilities of selection across subpopulations; (ii) participant non-response; (iii) and multiplicity (i.e. some patients were seen by multiple providers). We adjusted standard errors for the differential weighting and complex sample design using linearization methods. We used randomly drawn ‘donor’ responses within strata of respondents (‘hot deck’ imputation) to fill in the <5% of essential missing values. All analyses were conducted using Stata 7.0 (Stata Corp., College Station TX, USA).
|
Results |
|---|
|
TopMethodsResultsDiscussionReferences |
|---|
Nine hundred and twenty of the 2286 patients interviewed reported experiencing one of the three most bothersome symptoms during the previous 6 months. These 920 patients represented approximately 72779 HIV-infected persons in care in the USA, based on the HCSUS sampling design. The most frequent bothersome symptom, reported by 372 (40%) patients, was complicated cough, followed by diarrhea [332 (36%)], and weight loss [216 (24%)]. Patients who experienced any of the three most bothersome symptoms were more likely to have a low CD4 count (P < 0.001) and to have medical insurance, particularly Medicare (P = 0.003), than patients without bothersome symptoms (Table 2). However, the two groups of patients did not differ in regard to other sociodemographic or clinical factors.
|
The proportion of patients receiving all indicated care (called ‘performance rate’) was relatively low for all three types of symptoms, but varied considerably (Table 3), with about 41, 74, and 65% of patients with complicated cough, weight loss, and diarrhea, respectively, receiving all indicated care. The combined performance rate for the three types of symptoms was 57%. Performance rates also varied by indicator type. For example, the performance rates for history and physical exam items were 87 and 97%, respectively, but only 48% for processes of care involving an invasive diagnostic tests such as an arterial blood gas, bronchoscopy, or endoscopy.
|
Unadjusted comparisons of performance rates for different patient subgroups by symptom type and all symptoms combined revealed few significant differences (Table 4). Only CD4 count showed a consistent pattern in which the performance rate tended to be higher for patients with a CD4 count <500 cells/microliter, although these differences were only significant when performance for all three symptom types was combined. The combined performance rate for all indicated care for patients with the lowest CD4 counts (i.e.count <50 cells/microliter) was still relatively low, at 60%. Uninsured patients tended to have lower performance rates for complicated cough and diarrhea, and Medicare patients tended to have higher performance rates for each of the three symptom types than patients with other types of insurance. However, as with CD4 counts, these relationships were only significant for the three symptom types combined. Hispanic patients tended to receive lower rates of indicated care than whites and blacks for all three symptoms, but this relationship was not significant. Other sociodemographic and clinical factors were not associated with performance rates.
|
Adjusted analyses revealed similar results (Table 5). Patients with CD4 counts >50 cells/microliter were less likely to receive all indicated care than patients with CD4 counts >50 cells/microliter. This pattern was seen for all three symptoms, although it was only significant for the combined symptom measure. Compared with privately insured patients, uninsured patients were less likely and Medicare patients were more likely to receive all indicated care, but these relationships only had borderline significance for the combined symptom measure. Having a provider with a very low HIV patient load showed significant effects, but these were inconsistent across symptom type and based on unstable estimates for the complicated cough, for which only 10 patients had a provider with a very low patient load.
|
|
Discussion |
|---|
|
TopMethodsResultsDiscussionReferences |
|---|
In this study we used a symptom-based framework to examine the extent to which patients in a nationally representative sample of adults receiving care for HIV disease in 1996 and 1997 received indicated care for three types of common, bothersome symptoms. We found that performance was less than optimal. Although performance was better for patients with more advanced disease, these rates were hardly ideal. Even among patients with CD4 counts <50 cells/microliter, only half of the patients presenting with complicated cough and less than two-thirds of the patients with diarrhea received all indicated care. Although low, these rates are consistent with performance rates on other quality indicators for HIV and other conditions [15,20]. These results are also consistent with performance rates in other analyses using symptom-based quality measures of HIV care [12,13].
Although clinical care often involves assessment and treatment of specific symptoms, most conventional indicators of quality focus either on access to services or on processes and outcomes of care for a given condition, without necessarily taking into account specific symptoms or their degree of burden on the patient. Symptom-based quality measures derived from patient reports may offer a useful supplement that captures domains which are distinct from those captured by other approaches to quality assessment.
Patients’ reports about their symptoms are reliable, and a better predictor of outcomes than physicians’ reports about patients’ symptoms [21,22]. Patients’ reports about care also have proven to be a reasonable approach to measuring whether a variety of processes of care occurred [23,24]. In addition, using quality indicators from a symptom-based framework and patient reports has at least two theoretical advantages. Firstly, these indicators tend to be more patient-centered than conventional condition-based measures. By focusing on symptoms that are particularly burdensome for a given patient, they direct attention to aspects of the disease and care that are most salient to the patient [21,25]. Addressing such symptoms can not only improve patient satisfaction, but also markedly enhance health-related quality of life for these patients [7,10,11,26]. Secondly, symptom-based indicators may more accurately reflect the way clinical practice and decision-making occurs. In many instances when patients present with a specific symptom, the underlying cause is not immediately known. Although mild, non-specific symptoms may not warrant specific actions by clinicians other than reassurance and watchful waiting, bothersome symptoms that may indicate a serious complication or worsening of an existing chronic illness generally do.
Performance rates in our study were better for patients with more advanced HIV disease and worse for patients without insurance. However, other patient characteristics typically associated with quality of care, such as African American race or low education, did not have any significant effects on care. The lack of an effect of patient social characteristics is consistent with the literature on clinical uncertainty and decision making. Many aspects of medical care, including decisions about evaluating and treating a patient with a given symptom, are inherently uncertain processes [27]. It is in such uncertain areas of care that practice variation may be greatest and non-clinical factors, including patient, physician, organizational, and economic factors, may exert their greatest influence [28]. Conversely, when there is clear risk to the patient if they do not receive a basic work-up or selected treatments, then the ‘zone of discretion’ for physicians’ decision-making tends to be small and social characteristics of patients has less impact [29]. In this study, we included only patients who were already known to be HIV positive and were presenting with moderately or severely bothersome symptoms that could indicate a serious complication or progression of their disease. Indeed, this subset of HCSUS patients tended to have more severe HIV disease than other patients from the baseline survey. Thus, it is not entirely surprising that non-clinical factors such as race and ethnicity had only small effects. The lack of strong associations between patient sociodemographic characteristics and care is also consistent with several other studies of HIV patients and other patients with serious conditions [13,30–33].
Consensus between patients and providers about expectations for care may also explain differences in results using a symptom-based versus conventional quality measures. Patient expectations about and demand for specific types of care are strongly influenced by factors such as patient perceptions of the intensity and perceived seriousness of their symptoms, their vulnerability to illness, and their experience with their conditions [34,35]. By limiting our analytic sample to HIV patients who reported moderate to severe symptoms, we may have ensured higher rates of patient–provider agreement about the desirability of diagnostic tests such as the rate of checking lipid levels in diabetics.
Consensus between providers about what constituted appropriate care for the specific symptoms considered in this study may be a more complex issue. Although we convened a panel of physicians with expertise in HIV care to identify expected processes of care for common symptoms, there were no widely used guidelines available at the time of the study, and the measures used in this study have not been previously validated. Thus, our measures may not represent a consensus among community HIV providers or may represent an unreasonably high standard of care. If so, the finding that relatively few patients received ‘all indicated care’ does not necessarily reflect poor quality per se, but rather the lack of clear standards or an unfair measure of quality. While we acknowledge this possibility, we are skeptical of this explanation. The specific quality indicators we used were not chosen to create a comprehensive or ‘state of the art’ measure of optimal care. Instead, only basic processes of care that a panel of HIV experts from different settings agreed represented the minimum processes of care they would expect a physician to do for a given bothersome symptom were included. Indeed, a variety of possible measures were not included, such as certain laboratory tests that could not be reliably reported by patients and which might lead to inaccurate estimates of quality. In addition, in a study examining timeliness of care for HIV symptoms that based standards of care on a national survey of physicians, in which strong consensus was present, actual performance rates were still quite low [13].
Because we only included indicators representing the minimum expected care, we required providers to perform all indicators for a given symptom, and did not give partial credit if they only performed some of the indicators. While reasonable, this conservative approach may underestimate the level of care being provided. For example, a provider who performed four out of the five indicated processes of care may have been providing better care than a provider who performed only one of the indicators, but in this study both were considered to have performed equally poor care. A related issue is that indicators for one symptom may have been more complex to perform than another. For example, providers may be more reluctant to perform a somewhat invasive test like an arterial blood gas than a lung exam, or to simply ask a patient about loss of appetite. To address this, we designed certain indicators to be conditional such that they were only required in specific circumstances. For example, an arterial blood gas for patients with cough was only required if they were also complaining of shortness of breath. If a patient was not short of breath, that indicator was not included in the calculation of whether all expected processes of care for a complicated cough were performed. Despite these provisions, we observed wide variation in performance rates by indicator type, ranging from 48% for invasive tests to 97% for indicators involving physical exam. More research is needed to understand the ways in which invasiveness and other characteristics of an indicated process of care affect performance of indicated care.
This study has several other limitations. Data were collected in 1996 and 1997. During the intervening period, HIV care standards have changed greatly. It is possible that results would differ if we developed an updated set of symptom-based indicators and applied it to a more recent sample of HIV patients. Moreover, the evidence base for many of the indicators was sparse, leading our expert panel to rely on their clinical experience in their construction. In addition, for patients with extremely advanced disease diagnostic tests or other processes of care may not have been performed due to physicians’ sense of futility or patient preference. If so, performance scores for this subset of patients would have been underestimated, accentuating the trend toward higher performance in patients with low CD4 counts. However, we suspect the number of such patients in our sample was small and estimates of overall performance rates were unlikely to be significantly biased.
We measured how bothersome a symptom was to patients, but we did not know whether they reported that a symptom was bothersome because of its severity, duration, or both. Thus, it is possible that in some cases, such as when a symptom was severe enough to be bothersome but only lasted a few days, performance of diagnostic tests or treatment may not have been warranted, and quality of care may be underestimated. Since we did not interview providers, it is also possible that providers were unaware of a patient’s symptoms, either because the patient did not tell them about it or because the provider failed to hear what their patient was telling them. Research has repeatedly demonstrated that HIV providers are often uninformed about their patients’ symptoms [21,22]. While one might reasonably argue that a provider has a responsibility to learn about their patient’s symptoms, it may not be reasonable to attribute incomplete management of unreported symptoms to poor quality of care. A related concern is that the likelihood of a patient successfully communicating the presence of a bothersome symptom to their provider may be associated with education (socioeconomic status) and race, and this may have affected our results. However, since performance was not associated with education or race, we suspect that the size of this type of bias was small. Recall bias about presence of symptoms or care received could have affected results and in some cases may have led to underestimates of the quality of care delivered. If recall varied by race or socioeconomic status, this also could have biased results.
In addition, we could not distinguish whether care was received in an outpatient clinic, as opposed to an emergency room or hospital, which may represent a lack of regular access to care for vulnerable groups. However, if access was a common problem for these patients with HIV and burdensome symptoms then we would have expected to find stronger associations between patient characteristics such as low levels of education or black race and quality, because these patient characteristics were associated with access to HIV care in other studies [15]. We chose our target symptoms based on our judgments of clinical importance in HIV care, and our conclusions may not apply to other symptoms and conditions. Finally, as the survey asked about all care for the most bothersome symptom in the past 6months, we could not distinguish between patients with a single episode and those with multiple episodes of care. It is possible that the number of episodes, or the inability to recall indicated care for more remote episodes, could affect our results.
In conclusion, using a symptom-based framework of quality, we found that HIV care for several common, burdensome symptoms was suboptimal. Although performance was better among subgroups of patients with low CD4 counts, a substantial number of these patients still did not receive indicated care. Quality of care for HIV and other conditions has been shown to improve as standards of care become more widely accepted and quality measures are routinely monitored [15,20]. This study suggests that further development and use of symptom-based guidelines for care and quality measures along these lines are warranted, and could help improve care in ways that are not well captured by more conventional quality measures.
This analysis was supported by the California Health Care Foundation. The overall HIV Cost and Services Utilization Study is being conducted under cooperative agreement U-01HS08578 (M.F.S., PI; S.A.B., Co-PI) between RAND and the Agency for Healthcare Research and Quality (AHRQ, formally the Agency for Health Care Policy and Research). Substantial additional funding for this cooperative agreement was provided by the Health Resources and Services Administration, the National Institute of Mental Health, the National Institute on Drug Abuse, and the National Institutes of Health Office of Research on Minority Health through the National Institute of Dental Research. Additional support was provided by VA HSR&D Career Development Award (S.A. and A.G.), the Robert Wood Johnson Foundation Grant Program, and Generalist Physicians Faculty Scholar Program, Merck and Company, Glaxo-Welcome, Inc., the National Institute on Aging, and the Office of the Assistant Secretary for Planning and Evaluation in the US Department of Health and Human Services.
|
References |
|---|
|
TopMethodsResultsDiscussionReferences |
|---|
- Mathews WC, McCutchan JA, Asch S et al. National estimates of HIV-related symptom prevalence from the HIV Cost and Services Utilization Study. Med Care 2000; 38: 750–762.[CrossRef][Web of Science][Medline]
- Kroenke K, Harris L. Symptoms research: a fertile field. Ann Intern Med 2001; 134: 801–802.
[Free Full Text] - Kroenke K. Studying symptoms: sampling and measurement issues. Ann Intern Med 2001; 134: 844–853.
[Abstract/Free Full Text] - Lorenz KA, Shapiro MF, Asch SM, Bozzette SA, Hays RD. Associations of symptoms and health-related quality of life: findings from a national study of persons with HIV infection. Ann Intern Med 2001; 134: 854–860.
[Abstract/Free Full Text] - Cleary PD, Fowler FJ Jr, Weissman J et al. Health-related quality of life in persons with acquired immune deficiency syndrome. Med Care 1993; 31: 569–580.[CrossRef][Web of Science][Medline]
- Cunningham WE, Rana HM, Shapiro MF, Hays RD. Reliability and validity of self-report CD4 counts in persons hospitalized with HIV disease. J Clin Epidemiol 1997; 50: 829–835.[CrossRef][Web of Science][Medline]
- Cunningham WE, Hays RD, Ettl MK et al. The prospective effect of access to medical care on health-related quality-of-life outcomes in patients with symptomatic HIV disease. Med Care 1998; 36: 295–306.[CrossRef][Web of Science][Medline]
- Hays RD, Cunningham WE, Sherbourne CD et al. Health-related quality of life in patients with human immunodeficiency virus infection in the United States: results from the HIV Cost and Services Utilization Study. Am J Med 2000; 108: 714–722.[CrossRef][Web of Science][Medline]
- Wu AW, Hays RD, Kelly S, Malitz F, Bozzette SA. Applications of the Medical Outcomes Study health-related quality of life measures in HIV/AIDS. Qual Life Res 1997; 6: 531–554.[CrossRef][Web of Science][Medline]
- Wilson IB, Cleary PD. Linking clinical variables with health-related quality of life. A conceptual model of patient outcomes. J Am Med Assoc 1995; 273: 59–65.
[Abstract/Free Full Text] - Watson A, Samore MH, Wanke CA. Diarrhea and quality of life in ambulatory HIV-infected patients. Dig Dis Sci 1996; 41: 1794–1800.[CrossRef][Web of Science][Medline]
- Kilbourne AM, Andersen RM, Nakasono R et al. Development and application of a method to assess timeliness of medical care for HIV symptoms. Health Serv Outcomes Res Methodol 2001; 2: 101–115.[CrossRef]
- Kilbourne AM, Andersen RM, Asch S et al. Response to symptoms among a U.S. national probability sample of adults infected with human immunodeficiency virus. Med Care Res Rev 2002; 59: 36–58.
[Abstract/Free Full Text] - Cunningham WE, Markson LE, Andersen RM et al. Prevalence and predictors of highly active antiretroviral therapy use in patients with HIV infection in the united states. HCSUS Consortium. HIV Cost and Services Utilization. J Acquir Immune Defic Syndr 2000; 25: 115–123.[CrossRef][Web of Science][Medline]
- Shapiro MF, Morton SC, McCaffrey DF et al. Variations in the care of HIV-infected adults in the United States: results from the HIV Cost and Services Utilization Study. J Am Med Assoc 1999; 281: 2305–2315.
[Abstract/Free Full Text] - Frankel MR, Shapiro MF, Duan N et al. National probability samples in studies of low-prevalence diseases. Part II: Designing and implementing the HIV cost and services utilization study sample. Health Serv Res 1999; 34: 969–992.[Web of Science][Medline]
- Whalen CC, Antani M, Carey J, Landefeld CS. An index of symptoms for infection with human immunodeficiency virus: reliability and validity. J Clin Epidemiol 1994; 47: 537–546.[CrossRef][Web of Science][Medline]
- Wilson IB, Roubenoff R, Knox TA, Spiegelman D, Gorbach SL. Relation of lean body mass to health-related quality of life inpersons with HIV. J Acquir Immune Defic Syndr 2000; 24: 137–146.[Web of Science][Medline]
- Wheeler DA, Gibert CL, Launer CA et al. Weight loss as a predictor of survival and disease progression in HIV infection. Terry Beirn Community Programs for Clinical Research on AIDS. JAcquir Immune Defic Syndr Hum Retrovirol 1998; 18: 80–85.[Web of Science][Medline]
- National Center for Quality Assurance. The State of Managed Care Quality, 2001. Washington, DC: NCQA, 2001.
- Justice AC, Rabeneck L, Hays RD, Wu AW, Bozzette SA. Sensitivity, specificity, reliability, and clinical validity of provider-reported symptoms: a comparison with self-reported symptoms. Outcomes Committee of the AIDS Clinical Trials Group. J Acquir Immune Defic Syndr 1999; 21: 126–133.[Web of Science][Medline]
- Justice AC, Chang CH, Rabeneck L, Zackin R. Clinical importance of provider-reported HIV symptoms compared with patient-report. Med Care 2001; 39: 397–408.[CrossRef][Web of Science][Medline]
- Cleary PD, Edgman-Levitan S, Roberts M et al. Patients evaluate their hospital care: a national survey. Health Aff (Millwood) 1991; 10: 254–267.[Medline]
- Stange KC, Zyzanski SJ, Smith TF et al. How valid are medical records and patient questionnaires for physician profiling and health services research? A comparison with direct observation of patients visits. Med Care 1998; 36: 851–867.[CrossRef][Web of Science][Medline]
- Jackson JL, Kroenke K. The effect of unmet expectations among adults presenting with physical symptoms. Ann Intern Med 2001; 134: 889–897.
[Abstract/Free Full Text] - Cunningham WE, Shapiro MF, Hays RD et al. Constitutional symptoms and health-related quality of life in patients with symptomatic HIV disease. Am J Med 1998; 104: 129–136.[CrossRef][Web of Science][Medline]
- Eddy DM. Variations in physician practice: the role of uncertainty. Health Aff (Millwood) 1984; 3: 74–89.[CrossRef][Medline]
- Eisenberg JM. Doctors Decisions and the Cost of Medical Care. Ann Arbor (MI): Health Administration Press Perspectives, 1986.
- Wennberg JE. Small area analysis and the medical outcome problem. In: Bunker J, ed. Research Methodology: Strengthening Causal Interpreations of Nonexperimental Data. Tuscon (AZ): U.S. Department of Health and Human Services. 1990; pp 177–206.
- Flood AB, Fremont AM, Jin K, Bott DM, Ding J, Parker RC Jr. How do HMOs achieve savings? The effectiveness of one organization’s strategies. Health Serv Res 1998; 33: 79–99.[Web of Science][Medline]
- Stein MD, Cunningham WE, Nakazono T et al. Care of vaginal symptoms among HIV-infected women. J Acquir Immune Defic Syndr 2000; 25: 51–55.[CrossRef][Web of Science][Medline]
- Keeler EB, Rolph JE. How cost sharing reduced medical spending of participants in the health insurance experiment. J Am Med Assoc 1983; 249: 2220–2227.
[Abstract/Free Full Text] - Menke TJ, Rabeneck L, Hartigan PM, Simberkoff MS, Wray NP. Clinical and socioeconomic determinants of health care use among HIV-infected patients in the Department of Veterans Affairs. Inquiry 2000; 37: 61–74.[Web of Science][Medline]
- Kravitz RL. Measuring patients’ expectations and requests. Ann Intern Med 2001; 134: 881–888.
[Abstract/Free Full Text] - Kravitz RL, Callahan EJ, Paterniti D, Antonius D, Dunham M, Lewis CE. Prevalence and sources of patients’ unmet expectations for care. Ann Intern Med 1996; 125: 730–737.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. J Saleem, E. S Patterson, L. Militello, S. Anders, M. Falciglia, J. A Wissman, E. M Roth, and S. M Asch Impact of Clinical Reminder Redesign on Learnability, Efficiency, Usability, and Workload for Ambulatory Clinic Nurses JAMIA, September 1, 2007; 14(5): 632 - 640. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||