Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
CHAPTER 8 ASSOCIATION BETWEEN DIARRHEA AND QUALITY OF LIFE IN HIV INFECTED PATIENTS RECEIVING HIGHLY ACTIVE ANTIRETROVIRAL THERAPY A. Tramarin, N. Parise S. Campostrini D. Yin M.J. Postma R. Lyu R. Grisetti A. Capetti A.M.Cattelan M.T. Di Toro A. Mastroianni E. Pignattai V. Mondardini G. Calleri E. Raise F. Starace for the Palladio Study Group Published in Quality of Life Research (2004) 13:243-50 Abstract Diarrhea is a common symptom that many HIV patients experience either as a consequence of HIV-infection or of highly-active antiretroviral therapy (HAART). A multicenter, prospective observational study was conducted in 11 AIDS clinics in Italy to determine the effect of diarrhea on health-related quality of life among patients receiving HAART. The study enrolled 100 consecutive HIV positive patients who had diarrhea while on HAART. For each enrolled patient a control patient with matching disease stage who did not have diarrhea was identified using existing data from another prospective observational study conducted in 34 AIDS clinics (including the 11 in current study). Quality of life was measured by MOS-HIV Health Survey (MOS-HIV). Paired t-test and multiple regression analysis were used to compare the quality of life among patients with and without diarrhea. Mean patient age was 40±7 years; 69% were male. Mean CD4 cell count was 342± 239 cells/mm3; 59% had AIDS. Of the cases, 49 patients had severe diarrhea (>5 bowel movements or >3 watery per day) and 46 patients had moderate diarrhea (3-5 bowel movements). Compared to matched control patients, cases experiencing diarrhea while on HAART had significantly lower MOS-HIV scores in all domains. The significant adverse effect of diarrhea on quality of life should be considered when choosing the appropriate antiretroviral drugs regimen. Introduction Next to viral suppression, improvement of quality of life is one of the primary objectives of antiretroviral therapy [1]. Since 1997, expert consensus guidelines have recommended the use of 2 nucleoside reverse transcriptase inhibitors (NRTIs) and a protease inhibitor (or 2 protease inhibitors) for the treatment of established HIV infection [2]. The combination of 2 NRTIs and a nonnucleoside reverse transcriptase inhibitor (NNRTI) has recently become part of the recommended regimens [1, 3]. The use of these highly active antiretroviral therapies (HAART) is associated with side effects that pose difficulty with adherence and pose a challenge to both physicians and patients to achieve the objectives of antiretroviral therapy, including quality-of-life improvement. Fat maldistribution, hyperglycemia/diabetes mellitus, hyperlipidemia have been reported in patients receiving HAART. Diarrhea is also a relatively common 179 side effect of some antiretroviral treatments. For example, the incidence of diarrhea associated with protease inhibitors ranges from 0 to 56% (33 to 56% for amprenavir; 14 to 32% for nelfinavir; 13 to 22% for ritonavir; 12 to 20% for saquinavir; and 0 to 4.6% for indinavir) [4]. An additional cause of diarrhea among HIV infected patients is opportunistic bowel infections (e.g. Salmonella, Cryptosporidium, Isospora, Cytomegalovirus). Bowel infections and HAART-associated diarrhea may have an acute or a chronic clinical course. Abdominal pain from acute diarrhea may be steady, colicky, intermittent or cramping with occasional exacerbation. Chronic diarrhea is less painful but has a greater impact on the daily life of patients. It is well known, for example, that in patients with irritable bowel the embarrassing nature of this symptom causes an emotional upset, which is experienced as major limit in social life [5]. Some regard HAART-associated diarrhea as a minor nuisance but its effect on patient quality of life is yet not well understood. Previous studies found that diarrhea and related illnesses had a negative effect on the quality of life of HIV-infected patients [6-7]. However, these studies were conducted in the early 90’s when HAART was not available. Diarrhea may have implications on quality of life and, in turn, on adherence. This may influence the prescribing of antiretroviral regimens. The aim of this study is to estimate the effect of diarrhea on health related quality of life in HIV-infected patients receiving HAART. Methods Study design This study used data from two cohort studies of HIV patients and was based on a case control study design. Cases were taken from a group of patients enrolled in a prospective observational study conducted in 11 AIDS clinics in Italy from June 2000 to February 2001. All adult patients experiencing diarrhea while on HAART were conceived as cases. Controls were taken from another prospective, observational study conducted in 34 AIDS clinics in Italy between 1998-99 (the Palladio study) [8]. The Palladio study also included those 11 clinics participating in the first study, from which cases were drawn. All patients enrolled in the Palladio study received a HAARTregimen including indinavir. For each case patient, a control patient without diarrhea and with at least matching CDC-disease stage was selected from 180 the control cohort. Patient quality of life was measured in both studies by Medical Outcome Study HIV Health Survey questionnaire (MOS-HIV) [ 9]. Patient-related data As indicated, cases were adult HIV-infected patients (aged 18 years or older) on HAART, visiting their physician and reporting diarrhea. Patients enrolled had diarrhea within 7 days of visit. Cases had to be under treatment with any combination including at least 3 antiretroviral agents in use in the AIDS clinics at February 2001. All selected controls had a triple combination including indinavir and the following NRTIs: zidovudine, lamivudine, stavudine, didanosine, zalcitabine. Demographic and clinical information such as CD4 cell counts and plasma viral loads were obtained from both studies. Cases and controls were classified into one of nine disease stages based on their clinical presentation (i.e., asymptomatic, symptomatic, AIDS defining events) and CD4 cell count (<=200 cells/mm3, 200-499 cells/mm3, >=500 cells/mm3). Severity of diarrhea was defined in four grades: Grade 1: 3 to 5 stools per day; Grade 2: 5 to 7 stools or 3 to 5 watery stools per day; Grade 3: more than 7 stools or 5 to 7 watery stools per day; and Grade 4: with more than 7 watery stools per day. Grade 1 diarrhea was considered moderate (mild or transient or increased looseness per day). From grade 2 to grade 4 diarrhea was conceived as severe. This classification was based on the assumption that intensity and duration of diarrhea associated abdominal pain, rarely has an impact on quality of life higher than fecal incontinence [6]. For this reason, the number of bowel movements was considered more important for classifying diarrhea. A number of 5 movements or watery stools per day was considered relevant for the risk of fecal incontinence and limitation to social life. Identification of matched controls The matched control patients were those that did not report specific adverse events potentially associated to HAART and/or generic symptoms inducing the physicians to stop the therapy within the 6 months (i.e. control patients were asymptomatic or had only negligible symptoms at the moment of enrollment). The primary matching criterion was disease stage. CD4-cell counts (within 10 cells/mm3) and the nearest age were used as additional 181 matching criteria if two or more patients matched according to the first criterion. MOS-HIV Cases and controls in both cohorts were invited to complete the MOS-HIV questionnaire. The MOS-HIV questionnaire was adapted from the SF-20 questionnaire by Wu et al. to address specific quality-of-life issues for persons with HIV disease by removing questions not relevant to HIVinfected persons and adding questions that assess relevant aspects of health, specifically for HIV-infected persons [9]. The MOS-HIV questionnaire measures 10 dimensions of health comprising general health perception, physical function, role function, social function, cognitive function, bodily pain, mental health, energy/fatigue, health distress, and quality of life. One further item assesses health transition. Dimensions of the MOS-HIV are scored on a 0-100 scale where higher scores indicate better health. The questionnaire has been validated and adapted in many countries and is widely used in clinical trials [10]. Statistics Paired t-test was used to compare MOS-HIV domain-specific quality of life scores (differences 6 months –baseline) between patients with diarrhea and their matched controls. All statistical tests were 2-tailed; results with a pvalue of < 0.05 were considered significant. To confirm the result obtained from the case control study design, a stepwise linear regression analysis has been performed. All diarrhea and control patients were taken together and the score in each dimension of MOS-HIV Health Survey at 6 months was assumed as dependent variable. The inclusion in the diarrhea group (yes/no) was considered as the independent variable of major interest. Age, CD4 count, plasma viral load, gender and CDC disease stage were considered as covariates. The order of including covariates into the model was based on each covariate’s individual effect on the squared multiple correlation coefficient (R2). 182 Results Patient Characteristics We enrolled 100 and 432 adult HIV-infected patients as cases and potential controls, respectively. Of these potential controls, 340 patients had a full 6 months follow up. Further, after the exclusion of patients who experienced adverse events while on HAART, 271 patients remained to be matched to the 100 cases. A matched control was identified for 96 of the 100 case patients. No appropriate matches were found for the remaining 4 patients. Table 1 describes the socio-demographic and clinical characteristics of cases and matched controls at enrollment. Patients with diarrhea (n = 96) Patients without diarrhea (n = 96) 40.2 ± 6.8 38.4 ± 7.5 68.8% 74.0% CD4 cell counts (cells/mm3, mean ± 1 standard deviation) 341.6 ± 239.0 335.7 ± 221.4 Plasma vRNA load (thousand copies/mm3, mean ± 1 standard deviation) 14 ± 89 14 ± 77 32 17 13 13 0 0 96 0 0 0 Patients Characteristics Age (mean years ± 1 standard deviation) Gender (percent male) Protease inhibitor regimen (number of patients) Nelfinavir Indinavir Ritonavir Saquinavir Amprenavir 183 NNRTI regimen (number of patients) Efavirenz Nevirapine Delavirdine 14 10 2 0 0 0 NRTI regimen (number of patients) Zidovudine Lamivudine Zidovudine + Lamivudine Stavudine Didanosine Zalcitabine Abacavir 18 59 18 54 13 18 12 37 70 15 52 15 9 0 Table 1. Demographic, clinical and therapeutic characteristics Patients with diarrhea and their matched controls had the same distribution over disease stage. Eighteen (18.7%) were in the stage A of CDC 93 classification (stage A1: 14; stage A2: 2; stage A3: 2); thirty-three (34.3%) were in stage B (stage B1:10; stage B2: 13; stage B3:10); and forty-five (46.9%) had AIDS (stage C1: 10; stage C2:8; stage C3: 27). They had similar age (mean 40 ± 7 vs. 38 ±7.5 years), CD4 cell count (342 ± 239 vs. 336± 221 cells/mm3), RNA viral load (14,000 ± 89,000 vs. 14,000 ± 77,000 copies/mL). About 70% of the patients were male in both groups. All patients received HAART-regimens including a combination of NRTIs and a protease inhibitor (or 2 protease inhibitors) or an NNRTI. Lamivudine and stavudine were the most commonly prescribed NRTIs in both groups. Among patients with diarrhea, prescribed protease inhibitors included nelfinavir (n=32), indinavir (n=17), ritonavir (n=13) and saquinavir (n=13). Four of these patients received both indinavir and ritonavir and 2 patients received ritonavir and nelfinavir. The NNRTIs prescribed to cases included efavirenz (n=14), nevirapine (n=10) and delavirdine (n=2). All control patients received indinavir as a part of their HAART-regimen. Among the 96 patients with diarrhea, about 51% had severe diarrhea (36 patients with grade 2, 6 with grade 3, 7 with grade 4) and 49% had mild diarrhea (grade 1). 184 Quality of Life Measures Compared to their matched controls, patients with diarrhea had significantly lower scores in all 11 MOS-HIV domains. Figure 1 shows the differences in average MOS-HIV domain scores with 95% confidence intervals for cases and controls. These differences ranged from 8 for domains, such as physical and cognitive functioning, to 19 for social functioning. These differences indicate that HIV-infected patients without diarrhea had significantly better quality of life compared to patients with diarrhea. Differences in MOS-HIV Scores 30 20 19* 18* 17* 15* 13* 11* 10 8* 17* 10* 16* 8* 0 GH PF RF PN SF MH EF HD CF QL HT MOS-HIV Domains Legenda: * p < 0.05, Paired t-test.; general health (GH), physical -function (PF), role-function (RF), social-function (SF), pain (PN), mental health (MH), energy/fatigue (EF), health distress (HD), cognitive- functioning (CF), quality of life (QL), health transition (HT). Figure 1. Differences in MOS-HIV domain scores between patients without diarrhea and those with diarrhea 185 We also compared the quality-of-life scores between the two groups by their disease stage. Among AIDS patients, statistically significant differences in MOS-HIV domain scores were found in 5 domains comprising general health, social functioning, energy/fatigue, quality of life, and health transition. Among HIV-positive patients, patients with diarrhea had significantly lower scores in all domains (Figure 2). Better QoL 100 90 MOS-HIV Scores 80 70 60 50 40 30 Worse QoL 20 GH PF RF PN SF MH EF MOS-HIV Domains HD CF QL HT HIV patients with diarrhea (n = 39) AIDS patients with diarrhea (n = 57) HIV patients without diarrhea (n = 39) AIDS patients without diarrhea (n = 57) Legenda: general health (GH), physical -function (PF), role-function (RF), social-function (SF), pain (PN), mental health (MH), energy/fatigue (EF), health distress (HD), cognitivefunctioning (CF), quality of life (QL), health transition (HT). Figure 2. HIV/AIDS disease stage - specific MOS-HIV domain scores 186 For patients with diarrhea, we also explored the association between quality-of-life score and the severity of diarrhea. Difference was observed in 2 domains where patients with severe diarrhea had lower scores in role and social functioning compared to patients with mild diarrhea (Figure 3 ). Better QoL 100 MOS-HIV Scores 90 80 70 60 50 40 30 Worse QoL 20 GH PF RF PN SF MH EF HD CF QL HT MOS-HIV Domains Patients with moderate diarrhea (n = 46) Patients with severe diarrhea (n = 50) Legenda: general health (GH), physical -function (PF), role-function (RF), social-function (SF), pain (PN), mental health (MH), energy/fatigue (EF), health distress (HD), cognitivefunctioning (CF), quality of life (QL), health transition (HT). Figure 3. MOS-HIV domain scores of HIV infected patients with moderate and severe diarrhea Table 2 shows the regression analysis coefficients for each MOS-HIV domain for cases (n=100) and those patients from the control cohort with full information available (n=340). The results of the regression analysis are consistent with those of the case-control analysis. The presence of diarrhea is always significant with the exception of the domain score for cognitive function (p = 0,444). 187 MOS-HIV Dimensions General Health (GH) Physical Function (PF) Role Function (RF) Pain (PN) Social Function (SF) Mental Health (MH) Energy/Fatigue (E/F) Health Distress (HD) Cognitive Functioning (CF) Quality of Life (QL) β diarrhea (stepwise) -28.7 -10.5 -13.6 -13.5 -17.3 -18.3 -26.1 -14.0 -4.2 -31.4 p-value R2 ** .000 .047 .009 .012 .001 .001 .000 .010 .444* .000 .166 .110 .149 .058 .152 .093 .177 .073 .017 .139 Legenda: * p > 0.05; ** full specifications of estimated functions: GH= -.230*stage -.287*diarrhea.; PF = -.289*stage -.105*diarrhea; RF= -.268*stage -.157*viral load -.136*diarrhea; PN = -.169*stage -.135*diarrhea; SF = -.208*stage-.108*viral load -.124*gender -.117*age -.173*diarrhea; MH = -.119*stage -.153*viral load; EF =-.261*diarrhea -.184*stage -.175*viral load.183*diarrhea; HD = -.154*stage -.109*viral load-.140*diarrhea; CF = -.128*stage; QL = -.160*viral load -.314*diarrhea. Table 2. Results from the regression analysis (regression coefficients and their significance). Discussion Viral suppression is the primary goal of antiretroviral therapy in HIV. However, improving patient quality of life is a further important objective, the more as a better quality of life may also influence patients' adherence to antiretroviral therapy. Using observational data, this study examined the relationship between diarrhea and HAART and its detrimental effect on quality of life. Each patient was asked to fill out the MOS-HIV questionnaire in the time periods considered and a case-control approach was used, matching by disease severity, CD4 cell count and age. By controlling for these major characteristics it was possible to assess the likely impact of diarrhea on quality of life in HIV-patients receiving HAART. We observed significant differences in quality-of-life outcomes in all MOS188 HIV domains between the case patients and matched controls. Whilst HIVpositive patients with diarrhea had significantly lower score in all domains, AIDS-patients had statistically significant differences in MOS-HIV domain scores in only 5 domains. These results are consistent with those published by other investigators, who also reported different impacts on quality-of-life outcomes between asymptomatic and AIDS patients after the start of effective antiretroviral therapy [11]. The role of HAART-related toxicity was considered a sufficient explanation for decreases in quality of life in asymptomatic HIV patients. In contrast, the overall positive outcome of HAART on quality of life is so high in AIDS patients to justify any concomitant negative effects of HAART-associated toxicity. In this study, severity of diarrhea was found to be associated with poor quality of life of HIV-patients. A statistical significant difference was found between patients with moderate and severe diarrhea in 2 domains of MOSHIV: social function and role function. A similar classification of diarrhea severity was used by another study [6]. It concluded that diarrhea was considered severe by patients over the cutoff point of 5 bowel movements per day. The presence (or the possibility) of fecal incontinence was considered as the major factor limiting social life and, in turn, affecting quality of life. A confirmation of this conclusion has also been reported in studies examining the impact of diarrhea on quality of life in elderly with gastrointestinal disease [12]. The infection with HIV results in a chronic disease. There is a decline in functional status and many manifestations and symptoms that may compromise the sense of well being. Cunningham examined the impact of constitutional symptoms (weight loss, mialgyas, fever, night sweats, diarrhea, fatigue/exhaustion) on quality of life of HIV-infected patients. Diarrhea, weight loss and fever were the symptoms associated with worse scores of quality of life and with progression of the disease [13]. In an earlier study, Watchel et al. found that neurologic symptoms, constitutional symptoms, dyspnea and diarrhea were commonly present among patients with AIDS [14]. The presence of symptoms was the strongest indicator of lower MOS-HIV scores in these patients. Similarly, HIV-related symptoms were found to be strongly associated with physical and mental health in the HIV-Cost and Service Utilization Study [15]. Health related quality-of-life measures often include symptoms. In a recent paper, Lorenz examined the association of several symptoms with two single items global measures of health related quality of life (i.e. perceived health, perceived quality of life) 189 and an association of diarrhea with worse perceived quality of life was found [16]. Moderate diarrhea is often underreported by HIV patients to their physicians [17] . However, it has been documented that patients with chronic diarrhea have 50% higher charges resulting from more physician visits and diagnostic testing and have more disability days and absenteeism [18]. We note that, in general, use of HAART has reduced morbidity and mortality [19-21] and hospital service use [22-23] drastically, in developed countries. Recent reviews of studies that examined the cost-effectiveness of HAART regimens concluded that HAART was remarkably cost-effective for the treatment of HIV infection, despite its sometimes (costly) adverse effects [24,25] . Compared to the use of a randomized-controlled-trial design, in which one group of patients is on HAART and the other receives placebo, limitations of our observational design should be noted. In addition, we acknowledge that in identifying controls comorbidity was not formally taken into account (e.g. hepatitis comorbidity). Furthermore, control patients may not always have reported all symptoms to their physicians but, in this case, it would involve a potential underestimation of the investigated effect. Although patients are matched according to CDC disease stage, CD4 cell counts, and age, potential differences such as patient location, treatment history, socioeconomic status, and other intrinsic characteristics may still exist between patients who did and did not have diarrhea. However, current standards in HIV-treatment do not allow the set-up of a clinical trial as specified above. Therefore, current conclusions should often rely on observational investigations, such as ours. Furthermore, if the intervention effect remains sufficiently large after accounting for potential major confounding factors as we did, it is at least hypothesized that the symptom (diarrhea) does have an impact on the outcomes (MOS-HIV dimensions) being analyzed [24]. 190 Conclusion The list of possible etiologies of diarrhea is wide: it may be attributed to infectious agents, gastrointestinal malignancies, HIV enteropathy, idiopathic etiology. However, in the era of HAART, medications may be an important cause of diarrhea in HIV/AIDS- patients. We found that, diarrhea is associated with lower quality of life in this patient group. Therefore, careful consideration must be given when choosing antiretroviral medications for patients infected with HIV in order to reduce the potential incidence of avoidable drug-induced diarrhea, ensure better quality of life and enhance treatment adherence. 191 References 1. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. Panel on Clinical Practices for Treatment of HIV Infection convened by the Department of Health and Human Services (DHHS) and the Henry J. Kaiser Family Foundation. April 23, 2001. http://www.hivatis.org/guidelines/adult/Apr23_01/pdf/AAAPR23S.PDF 2. Carpenter CC, Fischl MA, Hammer SM, et al., Antiretroviral therapy for HIV infection in 1997. JAMA 1997;227:1962-1969. 3. Carpenter CC, Cooker DA, Fischl MA, et al., Antiretroviral therapy in adults: updated recommendations of the International AIDS SocietyUSA Panel. JAMA 2000;283:381-390. 4. Sherman DS and Fish DN. Management of protease inhibitor – associated diarrhea. Clin Infect Dis 2000;30:908-14. 5. El Serag HB, Olden K, Bjorkman D: Health – related quality of life among persons with irritable bowel syndrome: a systematic review. Aliment Pharmacol Ther 2002; Jun 16(6):1171 –1185. 6. Snijders F, de Boer JB, Steenbergen B, Schouten M, Danner SA, and van Dam FS. Impact of diarrhea and fecal incontinence on the daily life of HIV-infected patients. AIDS Care 1998;10:629-37. 7. Watson A, Samore MH, and Wanke CA. Diarrhea and quality of life in ambulatory HIV-infected patients. Dig Dis Sci 1996;41:1794-800. 8. Tramarin A, Alessi F, Calleri G and De Lalla F. The Palladio Pharmacoeconomic Study (PPES): Preliminary results. 37th of Annual Meeting Infectious Diseases Society of America, Philadelphia, 18-21 November 1999, Abstract book page 132. 9. Wu AW, Rubin HR, Mathews WC, et al., A health status questionnaire using 30 items from the Medical Outcome Study preliminary validation in person with early HIV infection. Med Care 1991;29:786-798. 10. Wu AW, Revicki DA, Jacobson D, Malitz FE: Evidence for reliability, validity and usefulness of the Medical Outcome Study HIV Health Survey (MOS-HIV). Qual Life Res 1997;29:786-798. 11. Bucciardini R, Wu AW, Floridia M, Fragola V, Ricciardulli D, Tomino C, Weimer LE, Pirillo MF, Mirra M, Marzi M, Giannini G, Galluzzo CM, Andreotti M, Massella M, Stefano Vella for the Istituto Superiore di Sanità 047 study: Quality of life outcomes of combination zidovudine-didanosine-nevirapine and zidovudine-didanosine for antiretroviral-naive advanced HIV-infected patients. AIDS 2000;14:2567-2574. 192 12. O’Keefe EA, Talley NJ, Zinsmeister AR, Jacobsen SJ: Bowel disorders impair functional status and quality of life in the elderly population – based study. J Gerontol A Biol Sci Med Sci. 1995; 50(4): 84-89. 13. Cunningham WE, Shapiro MF, Hays RD, Dixon WJ, Visscher BR, George WL, Ettl MK, Beck K. Constitutional Symptoms and Health – related Quality of Life in Patients with Symptomatic HIV Disease. Amer J Med 1998; 104:129-136. 14. Lubeck DP, Bennett CL, Mazonson PD, Fifer SK, and Fries JF. Quality of life and health service use among HIV-infected patients with chronic diarrhea. J Acquir Immune Defic Syndr 1993;6:478-84. 15. Wachtel T, Piette J, Mor V, Stein M, Fleishman J, Carpenter C. Quality of life in persons with human immunodeficiency virus infection: measurement by the Medical Outcomes Study instrument. Ann Intern Med 1992 Jan 15;116(2):129-37. 16. Hays RD, Cunningham WE, Sherbourne CD, Wilson IB, Wu AW, Cleary PD, McCaffrey DF, Fleishman JA, Crystal S, Collins R, Eggan F, Shapiro MF, Bozzette SA. 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; Jun 15, 08(9):714-22. 17. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med 1998;338:853-860. 18. Mocroft A, Vella S, Benfield TL, et al., Changing patterns of mortality across Europe in patients infected with HIV-1. EuroSIDA Study Group. Lancet 1998;352: 1725-1730. 19. Vittinghoff E, Scheer S, O'Malley P, et al., Combination antiretroviral therapy and recent declines in AIDS incidence and mortality. J Infect Dis 1999; 179:717-720. 20. Mouton Y, Alfandari S, Valette M, et al. Impact of protease inhibitors on AIDS-defining events and hospitalisations in 10 French AIDS reference centres. AIDS 1997, 11:F101-F105. 21. Keiser P, Kvanli MB, Turner D et al. Protease inhibitors therapy is associated with decreased HIV related health care costs in men treated at a Veteran Administration Hospital. J Acquir Immune Defic Syndr Hum Retrovirol 1999 20:28-33. 193 22. Beck EJ, Mandalia S, Williams I, et al. Decreased morbidity and use of hospital services in English HIV-infected individuals with increased uptake of antiretroviral therapy 1996-1997. AIDS 1999;13:2157-2164. 23. Moore RD. Cost-effectiveness of combination HIV therapy: 3 year later. Pharmacoeconomics 2000;17:325-330. 24. Bos J M, de Jong - van den Berg LTW, Postma MJ. Pharmacoeconomic evaluation of intensified antiretroviral treatment strategies in HIV/AIDS. Expert Review of Pharmacoeconomics & Outcomes Research, 2001;1:77-84, 25. Black N. Why we need observational studies to evaluate the effectiveness of health care. BMJ 1996; 312: 1215-1218. The members of the Palladio Study group are the followings: Alessandria: Azzini M, Mantia E; Asti: Biglino A, Degioanni M; Aviano: Tirelli U, Schioppa O; Belluno: Caruso G; Bologna1: Chiodo F, Borderi M, Donzelli C; Bologna2: Gritti F, Coronado O; Bolzano: Mian P, Paternoster C; Brescia: Carosi G, Torti C; Como: Vigevani GM, Pusterla L; Cremona: Carnevale G; Ferrara: Ghinelli F, Carradori S; Forlì: Cancellieri C; Genova1: Piersantelli N, Lorusso C; Genova2: Bassetti D; Genova3: Pagano G, Camera M; Legnago (VR) Parrinello A, Rovere P; Mantova: Scalzini A, Alessi F; Mestre: Borri A, Panese S; Milano1: Milazzo F, Rizzardini G; Milano2, Cargnel M, Grisetti R; Modena: Esposito R, Del Borgo C; Rovigo: Carretta M, Masiero G; Padova: Meneghetti F, Scanio F; Pavia: Filice G, Patruno S; Schio (VI): Marranconi F, Merighi M; Torino1: Soranzo ML, Macor A; Torino2: Di Perri G, Trieste: Mascioli M, Errera G; Treviso: Vaglia A, Zanatta A; Trento: Mian P, Paternoster C, Dorigoni N; Udine: Ciccone L, Panzolli L; Vicenza: De Lalla F; Venezia: Rosini G; Verona: Concia E, Dal Bravo P. 194