Over the last 2 years there have been dramatic changes in knowledge regarding the course of HIV-infection and the therapeutic options available for the treatment of HIV-infected patients[1-6]. In particular, combination antiretroviral therapy has proved to be effective in decreasing viral load, increasing or maintaining CD4 lymphocyte counts and preventing clinical events[7-10]. The use of potent antiretroviral therapies has resulted in remarkable declines in hospitalization rates, morbidity and mortality where drugs are available[3,11,12], and several international expert committees on the treatment of HIV-infection have recommended that the impact of extended treatment on quality of life should become a major consideration[3,5].
Health-Related Quality of Life (HRQoL) is increasingly incorporated as an outcome in clinical trials of treatments for HIV-infected patients[13-18]. Several disease-specific instruments have been developed to measure HRQoL in HIV-infected patients, including the MOS-HIV[19,20], MQOL-HIV, AIDS-HAQ[22,23], HAT-QoL[24,25], FAHI, HIV-QL31  and the HOPES[28,29]. The MOS-HIV was one of a family of instruments which developed out of the Medical Outcomes Study (MOS), and has become one of the most widely used measures for assessing HRQoL in HIV-infected patients. The MQOL-HIV was developed recently and includes socio-economic aspects of HRQoL that are specific to HIV-infected patients, which had never before been included in HRQoL questionnaires for these patients.
Given the proliferation of HRQoL instruments in numerous areas of clinical research, the comparison of the psychometric properties of different questionnaires is becoming increasingly important, because researchers need to know which is the most valid and useful. For use in clinical research, considerations of particular interest might include the time of questionnaire administration, the appropriate coverage of dimensions, the level of reliability, and the capacity to detect changes over time due to antiretroviral treatment. The purpose of the present study was to compare the feasibility and psychometric performance of two health-related quality of life instruments for use in clinical research, the MOS-HIV and the MQOL-HIV, which aim to assess quality of life in HIV-infected patients.
Between November 1997 and June 1998, the MOS-HIV and MQOL-HIV were administered randomly to a sample of 558 male and female HIV-infected patients between 18 and 65 years of age recruited in 23 Spanish hospitals. Subjects could be either treatment naive, or already receiving treatment. Patients were excluded if they were pregnant, were injecting drug users, had HIV acute infection, were unable to understand and answer the questionnaires, or were not willing to attend follow-up visits. Test-retest reliability was assessed in a sub-sample of 98 patients who were clinically stable with their current anti-retroviral treatment and who completed the MOS-HIV or MQOL-HIV questionnaire on two occasions separated by an interval of 2 weeks. Clinically stable patients were defined as those who had not experienced changes in health state or treatment during the 4 weeks prior to the study start date, and whose condition and treatment were not expected to change in the 2 weeks following their inclusion in the study. Sensitivity to change was assessed in a sub-sample of 296 patients initiating or changing anti-retroviral treatment due to lack of _efficacy, who completed the questionnaires at baseline and after 3 months.
Questionnaires were also administered to 80 healthy blood donors recruited from the Blood Transfusion Service of one of the participating hospitals. Healthy blood donors were recruited using age and sex quotas which were proportional to the Spanish population of registered AIDS cases. Questionnaire administration was randomized in both patients and healthy blood donors using a ‚block‚ design. Patient blocks consisted of three groups of four patients by centre, with each group of four patients representing a diagnostic category (asymptomatic, symptomatic and AIDS); within each block, the MOS-HIV-and MQOL questionnaires were assigned randomly, randomization having been carried out previously using the Sampsize Program version 2.0 (Blackwell Science Ltd, 1987). All participants provided written informed consent and the research protocol was approved by the Ethics Committees of the participating hospitals.
The MOS-HIV [19,20] consists of 35 questions which assess eleven dimensions of health including general health perceptions (GHP), pain (P), physical functioning (PF), role functioning (RF), social functioning (SF), mental health (MH), energy/fatigue (EF), health distress (HD), cognitive function (CF), quality of life (QL) and health transition (HT). All items are answered on a scale of two, three, five or six response options. MOS-HIV sub-scales are scored as summated rating scales and transformed linearly to a 0-100 scale where higher scores indicate better health. In addition to scores for each sub-scale, a physical health summary score (PHS) and a mental health summary score (MHS) are generated. These summary scores are obtained by standardizing the scores of the sub-scales using the weighting coefficients estimated in a previous study and provided by the author of the questionnaire (A. Wu, personal communication, 1998)[20,33].
The MQOL-HIV  consists of 40 questions which assess ten dimensions of health including mental health (MHE), physical health (PHE), physical functioning (PFU), social functioning (SFU), social support (SSU), cognitive functioning (CFU), financial status (FST), partner intimacy (PIN), sexual functioning (SXF) and medical care (MCA). All items are answered on a scale of seven or eight response options. MQOL-HIV _sub-scales are scored as summated rating scales on a 4-28 point scale, which in the present study were transformed to a 0-100 scale (where higher scores indicate better health) to be comparable with MOS-HIV sub-scale scores. Additionally, an overall index of quality of life is generated by means of the formula MQOL Index = (2 ∞ PHE score) + MHE score.
In order to test construct validity, all participants also completed the EuroQol 5-D (EQ-5D) [34,35] questionnaire. The EQ-5D is a generic questionnaire which measures health-related quality of life in five dimensions [mobility (M), self care (SC), usual activities (UA), pain/discomfort (PD) and anxiety/depression (AD)] each of which has three levels of severity (1_=_no problems, 2_=_moderate problems and 3_=_severe problems). The EQ-5D also uses a visual analogue scale (VAS) which is designed to look like a thermometer to measure overall, self-rated health. The VAS ranges from 0 (the worst health state imaginable) to 100 (the best health state imaginable).
The time frame used in all three HRQoL questionnaires was the previous 2 weeks. An additional item about changes in health status was included to assess patients‚ subjective perceptions of change in their health status between visits. The item was worded as follows: ‚How has your health status been since your last medical visit?‚ Response options were: much worse, worse, no change, better and much better.
The following socio-demographic variables were recorded: age, gender, educational level, job status and HIV-risk group. Clinical variables recorded included: chronic comorbid conditions, occurrence of infections and number of symptoms during the previous 2 weeks, time since diagnosis of HIV-infection, HIV infection stage, number of CD4, HIV-1 RNA levels in plasma, and antiretroviral therapy at baseline. Data on chronic comorbid conditions, infections and symptoms were obtained by the investigator during interviews with the patient. The HIV disease stage was designated by the investigator according to clinical criteria established by the Centers of Disease Control adapted for Europe. The number of CD4 cells and levels of HIV-1 RNA in plasma samples were determined by the standardized laboratory assays available in the study centres. Plasma viral load was quantified by quantitative reverse _transcription polymerase chain reaction (PCR) assay in nineteen hospitals (in 187 and 194 patients for the MOS-HIV and the MQOL-HIV respectively), and by the branched DNA (bDNA) and the NASBA assays in two other hospitals (60 and 8 patients, and 62 and seven patients in the MOS-HIV and MQOL-HIV groups, respectively).
The socio-demographic and clinical characteristics of patients receiving the MOS-HIV and MQOL-HIV were compared, as were the socio-demographic _characteristics of healthy donors receiving the two questionnaires, using the chi-squared test, Student‚s t test, analysis of variance and Pearson‚s product moment correlation where appropriate.
Feasibility was assessed by calculating the percentage of patients with at least one missing response on the questionnaires and the mean time of administration. Reliability was measured in terms of internal consistency and test-retest reliability. Internal consistency was analysed using Cronbach‚s agr; coefficient, and test-retest reliability was assessed using the Intraclass Correlation Coefficient (ICC). Test-retest reliability was analysed for the whole sample of patients initially included as stable and separately for those stable patients declaring no change on the item recording patients‚ subjective perceptions of change in their health state between the two questionnaire administrations. It is generally accepted that ICCs for measurement at group level should be 0.70 or greater.
Validity was analysed in terms of construct and discriminant validity. Construct validity was evaluated by analysing correlations between scores on each HIV questionnaire and CD4 cell count, viral load, number of symptoms and scores on the EQ-5D descriptive system and VAS. It was hypothesized that scores on the MOS-HIV and MQOL would correlate more strongly with number of symptoms and with EQ-5D scores than with CD4 cell count and viral load[19,21]. It was also expected that sub-scales measuring the same attribute on the disease-specific and generic questionnaires would correlate more strongly than sub-scales measuring distinct domains; for example, the anxiety/depression domain on the EQ-5D should correlate more strongly with the MH and HD domains on the MOS-HIV and the MHE score on the MQOL-HIV than with domains of physical functioning on either instrument. Correlations with CD4 cell count, viral load, number of symptoms and scores on the EQ-5D descriptive system and VAS were analysed using Spearman‚s correlation coefficient. The ability of the MOS-HIV and MQOL-HIV to discriminate between HIV-infected patients and healthy donors was assessed by calculating the area under the receiver‚s operating characteristics curve (AUC). Their ability to _discriminate between asymptomatic, symptomatic and AIDS patients, and between naive and pre-treated patients, was assessed using analysis of variance (with the Scheffe test), and Student‚s t test, respectively. It was expected that scores would be lowest in patients with AIDS, and highest in asymptomatic patients[38,39]. The influence of socio-demographic and treatment characteristics on HRQoL scores was assessed using the chi-squared test, Student‚s t test and analysis of variance (with the Scheffe test) where appropriate.
In order to evaluate the sensitivity to change of the MOS-HIV and MQOL-HIV, differences in patient scores before and after starting or switching anti-retroviral treatment were analysed using paired t tests and by calculating the standardized effect size (SES)[40,41]. Although the paired t test in principle requires normality of the variables, it is nevertheless relatively robust to violations of normality, and analysis showed that variables in the current analysis were sufficiently close to normality to ensure the appropriateness of using this test. The SES was chosen instead of the standardized response mean and Guyatt‚s index, as it is the statistic most commonly used to calculate effect size  and therefore allows comparison with other studies. The SES is obtained by calculating the difference between mean scores before and after changing treatment and dividing the difference by the mean standard deviation of the scores before changing treatment. Validity of change scores was analysed by correlating change scores on the MOS-HIV and MQOL-HIV with changes in CD4 cell count, viral load, number of symptoms and scores on the EQ-5D descriptive system and the _EQ-5D VAS.
For multi-item scales, mean substitution was used for missing items if ≤ 25% of the items on a sub-scale or the questionnaire were missing. Sub-scales or questionnaires with a higher proportion of missing responses were excluded from the analysis.
Sample size was calculated to be able to detect correlations of ε_0.2 between scores on the MOS-HIV or MQOL-HIV and other clinical variables, as well as permitting the detection of score differences between patients and healthy donors (with an expected _difference between these two groups of 16 points and a standard deviation of 21.8) with a significance level of 0.05 and a power of 80%. Where multiple _comparisons were performed, the significance level was set at P_<_0.01.
Of 558 patients included in the study, 275 completed the MOS-HIV and 283 completed the MQOL-HIV. Table 1 shows the socio-demographic and clinical characteristics of patients and socio-demographic characteristics of healthy donors. No statistically significant differences were observed between the two patient groups (MOS-HIV and MQOL-HIV) on socio-demographic and clinical variables. Statistically significant differences between patients and the control group were found for job status, with a higher proportion of healthy donors being in employment (P <_0.05).
A total of 18.9% of patients had at least one missing response on the MOS-HIV, with percentages of patients with at least one missing response ranging from 0.7% on the SF sub-scale to 6.9% on the MH sub-scale and 8.4% on the PF sub-scale. On the MQOL-HIV, 33.6% of patients had at least one missing response for the overall questionnaire, ranging by sub-scales from 1.8% on SSU to 9.9% on MHE and 11.7% on SXF. Mean time of administration for the MOS-HIV was 15.9 (SD_=_7.3) min and for the MQOL-HIV 16.3 (SD_=_7.9) min (P_>_0.05).
Table 2 shows Cronbach‚s agr; and ICC values for the MOS-HIV and MQOL-HIV. Cronbach‚s agr; values for the MOS-HIV sub-scales (0.78-0.89) were higher than those obtained for the MQOL-HIV (0.44-0.82). ICCs for stable patients for both questionnaires using the whole sample were satisfactory (≥_0.7) except for the P, PF, RF, SF, QL and HT sub-scales on the MOS-HIV and the MHE, PHE, PFU, CFU and FST sub-scales on the MQOL-HIV. When only patients declaring no change in their condition between the two administrations were analysed (71% of the reliability sub-sample on the MOS-HIV and 76% on the MQOL-HIV), the ICC coefficient for the PHS score and the P sub-scale on the MOS-HIV improved considerably from 0.58 to 0.76 and from 0.51 to 0.70, respectively, although for the remaining MOS-HIV and MQOL-HIV sub-scales there was no appreciable improvement in test-retest reliability.
In terms of construct validity, there were no statistically significant (P_<_0.01) correlations between CD4 cell count or viral load and any of the MOS-HIV and MQOL-HIV dimensions or summary scores. All MOS-HIV summary and dimension scores except HT correlated significantly (P_<_0.01) with number of symptoms, with correlations ranging from weak (0.28 for the HD dimension) to moderate (correlation of 0.49 for the PHS summary score). Seven of the 11 MQOL-HIV dimension and summary scores correlated significantly (P <_0.01) with number of symptoms. Correlations for the FS, SXF, SSU, FST, and PIN dimensions were not significant, whereas correlations for other dimensions were weak to moderate, ranging from 0.31 for the MH summary score to 0.42 for the MQOL Index and the SF dimension. Correlations between the two HIV questionnaires and EQ-5D dimensions and VAS ranged from weak to strong, although the MOS-HIV correlated more strongly and over a wider range of dimensions than the MQOL-HIV. Correlations were in the expected direction for both instruments, with, for example, the highest correlations for the EQ-5D VAS being with the GHP item on the MOS-HIV (0.66, P_<_0.01), and the highest correlations with the EQ-5D anxiety/depression dimension being with the MOS-HIV MHS (0.67), the MOS-HIV HD dimension (0.59) and the MQOL-HIV MH summary score (0.59).
Calculation of the AUC for both questionnaires showed that all the sub-scales of MOS-HIV except the HT sub-scale discriminated between patients and healthy donors (AUC from 0.6 to 0.86), as did all MQOL-HIV sub-scales except the SSU, CFU and PIN sub-scales. Table 3 shows that only the GHP, PF, RF and HT sub-scales for the MOS-HIV and the PHE, PFU, SFU, SXF and MCA sub-scales for the MQOL-HIV discriminated between asymptomatic and symptomatic or AIDS patients (P_<_0.05). Statistically significant differences between these groups of patients were found in age and job status, with asymptomatic patients being more likely to be younger and in employment (P_<_0.05). There were no statistically significant differences in HRQoL scores by age, but there were differences in HRQoL by job status. Treatment-naive patients reported poorer HRQoL than previously treated patients on almost all MOS-HIV sub-scales, but only on the CFU sub-scale on the MQOL-HIV _(P_<_0.05).
Table 4 shows effect sizes by dimension for both _questionnaires. Statistically significant improvements _(P_<_0.05) were found in over half the sub-scales in both questionnaires, although effect sizes in all sub-scales and summary scores were generally small. Correlations between MOS-HIV and MQOL-HIV change scores and changes on clinical variables were similar to those found when correlating MOS-HIV and MQOL-HIV scores with clinical variables (data not shown).
Very little research has been done on directly comparing HIV-HRQoL instruments for use in clinical research. Only one earlier study  compared the MOS-HIV with a self-administered version of an ad hoc HRQoL-HIV scale. This earlier, non-random study was performed in only 99 HIV-seropositive homosexual patients and showed that both scales had good internal consistency and good construct validity, with positive correlations with CD4 and CD8 cell count. The purpose of the present study was to compare the feasibility and psychometric performance of two HRQoL instruments in a large and clinically diverse sample of HIV-infected patients in the context of a randomized study.
The MOS-HIV appeared to be more feasible than the MQOL-HIV. Although administration time was similar, the MOS-HIV had a lower percentage of patients with at least one missing response. This may be partially due to the more intimate nature of some of the items in the MQOL-HIV, particularly in the SXF sub-scale, which had the highest percentage of patients with missing responses (11.7%). Additionally, a fairly high proportion of respondents (37.9%) did not have a partner, making it impossible to generate a summary score on the MQOL-HIV PIN sub-scale. Thus, although including the SXF and PIN dimensions may increase content validity, it also generates more missing responses, which is both inefficient and can lead to biased results if non-respondents differ systematically from respondents. With regard to content validity it is worth noting, however, that about 50% of domains covered are common to both questionnaires; of the remaining 50% the MQOL-HIV tends to focus on the impact of HIV on social aspects of QoL, whereas the MOS-HIV focuses more on health-related aspects of QoL. The implications of questionnaire content should be considered in the context of the particular objectives of the trial.
Even when only those patients reporting no change in health status were used to evaluate test-retest reliability, results were unsatisfactory for the PF, RF, SF, QL and HT sub-scales on the MOS-HIV, and for the MHE, PHE, PFU, CFU, and FST sub-scales on the MQOL-HIV. To some extent, this is a little surprising as many of these scales measure relatively stable concepts (such as physical and social functioning), at least in comparison with constructs which are usually considered more labile, such as mental health or health distress on the MOS-HIV, but which here obtained higher ICCs. On the MOS-HIV, this might be explained to some extent by large ceiling effects on some of the sub-scales which showed poor test-retest reliability (for example, the P, PF, RF and SF sub-scales had ceiling effects of 51.1, 24.4, 78.7 and 77.1%, respectively at baseline), as the ICC tends to decrease as ceiling effects increase. On the MQOL-HIV, however, the scales with the largest ceiling effects (35.2 and 34.7% on the SSU and PIN sub-scales, respectively), had acceptable ICCs, suggesting that other factors must have been responsible for the poor test-retest reliability, and it should be remembered that even in relatively stable patients with HIV or AIDS there may be considerable variability arising from uncertainty of the disease course  and compliance with anti-retroviral treatments.
In terms of construct validity, the MOS-HIV slightly outperformed the MQOL-HIV, with a greater number of MOS-HIV sub-scale scores correlating both with number of symptoms and with EQ-5D domains and VAS scores. Correlations were also generally slightly stronger on the MOS-HIV. Additional stepwise multiple regression analyses using MOS-HIV and MQOL-HIV scores as dependent variables and socio- demographic and clinical variables (including EQ-5D VAS scores) as independent variables failed to reveal combinations of clinical and/or socio-demographic variables which produced higher correlations with QoL variables, although some socio-demographic variables were associated with some MOS-HIV and MQOL-HIV sub-scales, for example, gender with the PHS, GHP, PF, RF and SF sub-scales on the MOS-HIV, and age with the MQOL-HIV MHE, SFU and SXF sub-scales. These results are consistent with those obtained in other studies using the same instruments, which found moderate to strong correlations with symptoms[19,21,45,46], but low correlations with CD4 cell count and HIV-1 RNA levels in plasma[21,45,47].
The MOS-HIV also discriminated more successfully between patients and non-HIV subjects, although_neither instrument was particularly successful at discriminating between asymptomatic, symptomatic, and AIDS patients. There was, nevertheless, a tendency for scores on both instruments to decrease as disease stage advanced. Unlike the MQOL-HIV, the MOS-HIV was capable of distinguishing between patients who were taking anti-retroviral treatment and those who were not. This may indicate that the failure to distinguish between disease stages was at least partially due to the more overriding influence of whether or not patients were taking anti-retroviral treatment, suggesting in turn that the advent of highly active anti-_retroviral therapy (HAART) makes the current HIV-infection classification system somewhat outdated, at least in terms of the impact of the disease on health status and quality of life.
Although the MOS-HIV was slightly more responsive than the MQOL-HIV to clinical changes in patients who started or switched anti-retroviral treatment, effect sizes were generally small for sub-scales on both questionnaires. Standardized effect sizes of over 0.2 were only found for the PF, EF, HD, CF and QL sub-scales on the MOS-HIV and the MHE, PHE and MCA sub-scales on the MQOL-HIV; results which are in agreement with those of earlier studies using the MOS-HIV[13,17,18,40]. In both questionnaires, effect sizes were lowest in dimensions reflecting social functioning (or interactions with others, as in the MQOL-HIV PIN and SXF sub-scales). Although this may be due to some extent to ceiling effects on the RF and SF MOS-HIV sub-scales, it may also be related to study duration, as changes in social functioning are perhaps unlikely to occur in a period as short as 3 months. Longer term studies would be useful in determining not only whether anti-retroviral treatment has any effect on social aspects of HRQoL, but also in determining how a longer term follow-up affected effect sizes in other dimensions. Responsiveness may also be affected by a number of factors, particularly the type of patient treated (naive or pre-treated), which could not be analysed here. A more detailed, stratified analysis is currently being carried out to determine whether effect sizes vary between different patient groups, or by _disease stage.
In conclusion, neither instrument demonstrated completely satisfactory psychometric properties. The results in terms of responsiveness were satisfactory given the type of treatment administered, although a longer _follow-up period would be helpful in determining the degree of change in domains which a priori might be expected to take longer to show change, such as social functioning. Although the MOS-HIV had fewer missing responses than the MQOL-HIV and showed greater construct and discriminant validity, it also had large ceiling effects on a number of sub-scales, which may limit their capacity to reflect change. Finally, given that almost half of the sub-scales on both instruments failed to achieve generally accepted standards of ICC coefficients of 0.70 and above, their use in clinical research in samples such as that studied here is clearly circumscribed.
We would like to thank A. Wu, N. Avis and K. Smith for their kind permission to adapt the MOS-HIV and MQOL-HIV questionnaires for use in a Spanish _population. We also greatly appreciate the help of the medical staff involved, particularly Dr. C. Ferran of the Bellvitge Hospital Blood Bank, and the participation of patients and healthy donors.
1. Mellors JW, Rinaldo CR, Gupta P, White RM, Todd JA, Kingsley LA. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science
2. Tashima K, Staszewski S, Morales-Ramirez J, et al. A_Phase III, multicenter, randomized, open-label study to compare the antiretroviral activity and tolerability of efavirenz (EFV) + indinavir (IDV) versus EFV + zidovudine (ZDV) + lamivudine (3TC) versus IDV + ZDV + 3TC at 36 weeks [Study 006]. 6th Conference on Retroviruses and Opportunistic Infections,
Chicago, 31 January-5 February 1999, [abstract LB-16].
3. Carpenter CCJ, Fischl MA, Hammer SM, et al
. Antiretroviral therapy for HIV Infection in 1998. Updated recommendations of the International AIDS Society-USA panel. JAMA
4. BHIVA Guidelines Co-ordinating committee. British HIV Association guidelines for antiretroviral treatment of HIV seropositive individuals. Lancet
5. Gazzard B, Moyle G. 1998 revision to the British HIV Association guidelines for antiretroviral treatment of HIV seropositive individuals. Lancet
6. Gatell JM, Clotet B, Podzamczer D, Miró JM, Mallolas J. Guía práctica del SIDA. Clínica, diagnóstico y tratamiento
. Barcelona: Masson, 1998.
7. Cameron DW, Heath-Chiozzi M, Danner S, et al
. Randomised placebo-controlled trial of ritonavir in advanced HIV-1 disease
8. Detels R, Muñoz A, McFarlane G, et al
. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. JAMA
9. Hogg RS, Rhone SA, Yip B, et al
. Antiviral effect of double and triple drug combinations amongst HIV-infected adults: lessons from the implementation of viral-load driven antiretroviral therapy. AIDS
10. Pialoux G, Raffi F, Brun-Vezinet F, et al
. A randomised trial of three maintenance regimens given after three months of induction therapy with zidovudine, lamivudine, and indinavir in previously untreated HIV-1-infected patients. N Engl J Med
11. Marschner IC, Collier AC, Coombs RW, et al
. Use of changes in plasma levels of Human Immunodeficiency Virus Type 1 RNA to assess the clinical benefit of antiretroviral therapy. J Infect Dis
12. Palella FJ Jr, Delaney KM, Moorman AC, et al
. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med
13. Wu AW, Rubin HR, Mathews WC, et al
. Functional status and well-being in a placebo-controlled trial of zidovudine in early symptomatic HIV infection. J Acquir Immune Defic Syndr
14. Safrin S, Finkelstein DM, Feinberg J, et al
. Comparison of three regimens for treatment of mild to moderate Pneumocystis carinii Pneumonia in patients with AIDS. Ann Intern Med
15. Cohen J. Statistical power analysis for the behavioral sciences.
2ª edn. rev. Hillsdale, New Jersey: Lawrence Erlbaum Associates Inc 1988.
16. Cunningham WE, Hays R, 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
17. Scott-Lennox JA, Mills RJ, Burt MS. Impact of zidovudine plus lamivudine or zalcitabine on health-related quality of life. Ann Pharmacother
18. Perry CM, Barman JA. Management of HIV infection: the potential role of the lamivudine/zidovudine combination formulation. Dis Manage Health Outcomes
19. Wu AW, Rubin HR, Mathews WC, et al
. A health status questionnaire using 30 items from the Medical Outcomes Study. preliminary validation in persons with early HIV infection. Med Care
20. Wu AW, Revicki DA, Jacobson D, Malitz FE. Evidence for reliability, validity and usefulness of the Medical Outcomes Study HIV Health Survey (MOS-HIV). Qual Life Res
21. Smith KW, Avis EN, Mayer KH, Swislow L. Use of the MQoL-HIV with asymptomatic HIV-positive patients. Qual Life Res
22. Lubeck DP, Fries JF. Changes in health status after one year for persons at-risk for and with HIV infection. Psychol Health
23. Lubeck DP, Fries JF. Assessment of quality of life in early stage HIV-infected persons: data from the AIDS Time-Oriented Health Outcome Study (ATHOS). Qual Life Res
24. Holmes WC, Shea JA. Performance of a new, HIV/AIDS-_targeted quality of life (HAT-QoL) instrument in asymptomatic seropositive individuals. Qual Life Res
25. Holmes WC, Shea JA. A new HIV/AIDS-Targeted Quality of Life (HAT-QoL) instrument: development, reliability, and validity. Med Care
26. Peterman AH, Cella D, Mo F, McCain N. Psychometric validation of the revised Functional Assessment of Human Immunodeficiency Virus Infection (FAHI) quality of life instrument. Qual Life Res
27. Leplège AP, Rude N, Ecosse E, Ceinos R, Dohin E, Pouchot J. Measuring quality of life from the point of view of HIV-positive subjects: the HIV-QL31. Qual Life Res
28. Schag AC, Ganz PA, Kahn B, Petersen L. Assessing the needs and quality of life patients with HIV infection: Development of the HIV Overview of Problems-Evaluation System (HOPES). Qual Life Res
29. Ganz PA, Schag AC, Kahn B, Petersen L. Assessing the quality of life of HIV infected persons: clinical and descriptive information from studies with the HOPES. Psychol Health
30. Tarlov AR, Ware JE, Greenfield S, Nelson EC, Perrin E, Zubkoff M. The Medical Outcomes Study. An application of methods for monitoring the results of medical care. JAMA
31. Burgess A, Dayer M, Catalan J, Hawkins D, Gazzard B. The reliability and validity of two HIV-specific health-related Quality-of-Life measures: a preliminary analysis. AIDS
32. Vigilancia epidemiológica del SIDA en España 1996. Informe trimestral n.º 1/1996.
Centro Nacional de Epidemiología, Instituto de Salud Carlos III, Secretaría del Plan Nacional sobre SIDA, Ministerio de Sanidad y Consumo.
33. Revicki DA, Sorensen S, Wu AW. Reliability and validity of physical and mental health summary scores from the Medical Outcomes Study HIV Health Survey. Med Care
34. Brooks R. EuroQol: the current state of play. Health Policy
35. Badia X, Schiaffino A, Alonso J, Herdman M. Using the EuroQol 5-D in the Catalan general population: feasibility and construct validity. Qual Life Res
36. Hays RD, Anderson R, Revicki D. Psychometric considerations in evaluating health-related quality of life measures. Qual Life Res
37. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology
38. Carretero MD, Burgess AP, Soler P, Soler M, Catalan J. Reliability and validity of an HIV-specific health-related quality-of-life measure for use with injecting drug users. AIDS
39. Revicki DA, Wu AW, Murray M. Change in clinical status, health status, and health utility outcomes in HIV-infected patients. Med Care
40. Cohen C, Revicki DA, Nabulsi A, Sarocco PW, Jiang P. and the Advanced HIV Disease Ritonavir Study Group. A randomised trial of the effect of ritonavir in maintaining quality of life in advanced HIV disease. AIDS
41. Kazis LE, Anderson JJ, Meenan RF. Effect sizes for interpreting changes in health status. Med Care
42. O‚Keefe EA, Wood R. The impact of human immunodeficiency virus (HIV) infection on quality of life in a multiracial South African population. Qual Life Res
43. Feinstein AR, Cicchetti DV. High agreement but low kappa: I. The problems of two paradoxes. J Clin Epidemiol
44. Allison PJ, Locker D, Feine JS. Quality of life: a dynamic construct. Soc Sci Med
45. Murri R, Ammassari A, Fantoni M, et al
. Disease-related factors associated with health-related quality of life in people with nonadvanced HIV disease assessed using an Italian version of the MOS-HIV Health Survey. J Acquir Immune Defic Syndr Hum Retrovirol
46. 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
47. Franchi D, Wenzel RP. Measuring health-related quality of life among patients infected with Human Immunodeficiency Virus. Clin Infect Dis
48. Hays RD, Anderson R, Revicki D. Psychometric considerations in evaluating health-related quality of life measures. Qual Life Res
The Spanish Group for Validation of the MOS-HIV and MQOL-HIV questionnaires
X Badia, M Garcia, M Herdman, Institut Universitari de Salut Pública de Catalunya (Barcelona); _D Podzamczer, E Consiglio, Ciutat Sanitària i Universitària de Bellvitge (Barcelona); C Miralles, Hospital Xeral (Vigo); V Asensi, Hospital Covadonga (Oviedo); JR Arribas, Hospital La Paz (Madrid); _M Causes, Hospital Carlos Haya (Málaga); K Aguirrebengoa, Hospital de Cruces (Bilbao); F Pulido, Hospital 12 de Octubre (Madrid); JA Iribarren, Hospital Ntra Sra de Aranzazu (San Sebastián); _E Pérez-Guzmán, Hospital Puerta del Mar (Cádiz); _MJ Pérez-Elías, Hospital Ramón y Cajal (Madrid); _M Riera, Hospital Son Dureta (Mallorca); MA Colmenero, Hospital Virgen de la Macarena (Sevilla); M Gorgolas, Fundación Jiménez-Díaz (Madrid); A Muñoz, Hospital Infanta Cristina (Badajoz); I Santos, Hospital La Princesa (Madrid); V Roca, Hospital Clínico San Carlos (Madrid); A Herrera, Hospital General (Valencia); V Bonora, Hospital La Fe (Valencia); D Dalmau, Hospital Mútua de Terrassa (Barcelona); J Barrios, Hospital de la Sta Cruz y San Pablo (Barcelona); S Echevarría, Hospital de Valdecilla (Santander); J Pedreira, Hospital Juan Canalejo (La Coruña); J Portilla, Hospital General Universitario (Alicante); C López-Lavid, R Úrbez, Merck Sharp & Dohme de España Laboratories, S.A. (Madrid). Cited Here...