People living with HIV and AIDS (PLWHA) undergo considerable deterioration in health-related quality of life (HRQL) during the course of the disease. Despite the effectiveness of highly active antiretroviral therapy (HAART), which has transformed HIV into a chronic condition for with access to treatment, living with HIV, for many, remains debilitating, and provokes high levels of psychological distress1 largely caused by fear of death and prevailing stigma.2 The quality of life for PLWHA is impacted physically, mentally, and socially, the 3 spectrums identified by the World Health Organization.3 PLWHA are also concerned with the kind of life they can expect to lead in the context of their infection.4
Health care providers and researchers increasingly use HRQL measures to identify the most effective and least harmful treatment combinations.5 It is worth noting that many HRQL measures used in HIV research were developed before the HAART era, including the MOS-HIV,6 FAHI,7 HOPES,8 HAT-QOL,9 AIDS-HAQ,10 and the MQOL-HIV.11 Therefore, these measures may not be attuned to the impact of contemporary treatments, sensitive enough to detect HRQL changes, or account for the long-term experience of living with HIV.4
Furthermore, the prevalence of HIV across continents and the increasing access to HAART in low- to middle-income countries (approximately 42%)12 necessitates that HRQL measures be applicable across diverse populations. However, most instruments in use were designed within single developed countries and translated into the respective languages when applied internationally. Therefore, such measures do not account for semantic differences between languages nor sociocultural context in terms of value systems, perceptions of health and health beliefs in their development. Consequently, arguably, such instruments may contain a systematic bias and lack the discrimination required to be valid across linguistic groups.
The need for a relevant contemporary and sensitive patient-reported outcome (PRO) instrument to measure HRQL in PLWHA populations is therefore pressing, not only for use in clinical trials but also in operational research.13 In this article, we describe the development of PROQOL-HIV (patient reported outcomes quality of life–HIV), an HRQL instrument designed to be HIV specific and multidimensional, sensitive to the impact of antiretroviral therapy, disease stage, and applicable across populations. It is derived from in-depth interviews with PLWHA, treated and untreated, conducted simultaneously in 9 countries, across 5 continents, and in 11 languages between 2007 and 2008.14
The development of PROQOL-HIV followed a 4-stage process: (1) development of a conceptual model and item generation; (2) development of source questionnaires; (3) linguistic validation; (4) psychometric validation. Figure 1 summarizes the process.
Stage 1: Conceptual Model and Item Generation
The first stage defined a conceptual model of HIV HRQL and generated items to reflect this model. Through literature review, expert consultation, and particularly through preliminary interviews with patients, primary HRQL themes that comprehensively reflected the day-to-day experience of living with HIV were established. Banks of potential items for a draft PROQOL instrument were drawn from interview verbatim.
Interview Guide Development
We used standard research procedures to search the literature published in the PubMed database from 1981 to 2007 to identify validated HIV HRQL instruments and studies that had used those instruments, as well as qualitative studies of HIV HRQL–related issues. Qualitative studies were reviewed and compared with quantitative research to determine which HRQL topics were covered by existing measures and which were not. Fifteen HIV experts in 11 collaborating research centers in 9 participating countries were consulted to identify additional HIV HRQL topics not covered by the existing measures.
An interview guide was developed to explore the existing and the newly identified themes. This comprised 107 questions, and began with 5 open-ended questions in which participants' would describe the impact of HIV on daily activity and quality of life. A total of 102 semistructured questions followed, which were broadly categorized into the following topics: self and body care, daily activities, physical activities, health perception, energy/fatigue, cognitive functioning, social relationships, emotions, and treatment.
The study was approved by the institutional review boards of the sites participating, and informed consent was obtained from all participants. A total of 152 interviews were conducted with PLWHA over 6 months in 2007. Between 10 and 20 interviews were conducted in each of the 11 centers across 9 countries in 11 local languages: Australian English (Australia), Portuguese (Brazil), Khmer (Cambodia), Mandarin and Cantonese (China), French (France), Marathi (India), Senegalese French and Wolof (Senegal), Thai, and US English (United States). Patients were selected to represent a range of demographic and clinical characteristics and potential HRQL issues within local populations. Countries and centers were chosen to represent a broad range of cultures and linguistic families (Indo-European, Sino-Tibetan, Austro-Asiatic, Tai-Kadai, and Niger-Congolese), health care systems and access to treatment, prevalence and incidence rates of HIV infection, living standards, and income levels.
A trained interviewer briefed participants on the objectives of the research and assured them of confidentiality. Interview duration was between 1 and 2 hours and they were recorded and transcribed. Sociodemographic and HIV-specific treatment and clinical data were collected as well. At the end of each interview, participants were asked to identify the 5 issues of most personal relevance.
Interview Data: Reduction and Thematic Analysis
Interviews were typed verbatim in 11 languages. Transcripts were then translated by TransPerfect into English or French, according to the linguistic heritage of the country. The researchers examined and extracted the most frequent and the most important statements according to patient ratings, and the most informative responses according to the investigators' knowledge. Through this process, interview data were reduced from 3375 to 490 pages.
Thematic analysis of the reduced interview data involved identification of recurring themes and concepts, and attention to commonalities and variation within emerging themes. An iterative process was employed, whereby interview data were reexamined and refined by the investigators and English- and French-speaking expert panels as new themes and concepts arose.15 Across countries, information saturation for common themes occurred within 7 interview analyses, and around 10 for rarer issues. As shown in Figure 2, 18 empirically grounded themes emerged, which then were parceled into 11 broader underlying themes.
Use of Verbatim Text for Item Generation
The raw statements extracted during interview data reduction were converted into grammatically and syntactically correct items to form banks of potential items; 228 in English and 214 in French. Items were distributed across the 11 themes.
Stage 2: Development of Source Questionnaires
The second stage involved item review and selection, and the development of source questionnaires in US English and French. Each bank of potential items was submitted to a panel of 6 English- and 9 French-speaking HIV experts comprising physicians, psychologists, nurses, and members of patient associations. Items were reviewed on 4 criteria: (1) relevance of item to theme, (2) ease of understanding and translatability, (3) potential sensitivity to change, and (4) preference for retention or deletion of each item. Based on majority agreement, 70 items were retained to form draft PROQOL questionnaires in English and French.
The 2 draft questionnaires were compared and harmonized with the help of professional translators. After harmonization, the English version was back-translated into French, and the French version back-translated into English. Draft and target versions were compared and discrepancies resolved to establish the source PROQOL questionnaires.
Stage 3: Linguistic Validation
The third stage involved linguistic validation of the source US English questionnaire in 6 target languages by TransPerfect. Following published methods,16–18 the US English 70-item source questionnaire underwent 2 forward translations and reconciliation, then backward translation and reconciliation into Portuguese (Brazil), Khmer (Cambodia), Mandarin (China), Marathi (India), and Thai and Wolof (Senegal). Cognitive debriefing was conducted with 5 native speakers of each language to test item clarity and comprehension. Items were revised where appropriate. Harmonization of all items across all versions was completed by an international meeting of the project translators. An Australian version was adapted from the US version and also included in the international harmonization process as well. These 7 languages, in addition to US English and French, encompassed the populations targeted for psychometric validation of the instrument, the fourth stage of the project.
During the translation and validation process, semantic issues were encountered with 12 of the 70 items. For instance, in Wolof (Senegal), there were no equivalent words for “stressed” or “depressed.” Translators opted for common Wolof-specific expressions that also appeared in interviews, which convey the meaning of stress (“a weight on my shoulders”) and depression (“less taste for life”) to maintain equivalent concepts across languages. In Mandarin, Marathi, and Khmer, a word meaning “future” replaced “long term” in the English translation, there being no direct equivalent. However, no irreconcilable translation issues were encountered.
Stage 4: Psychometric Validation
Stage 4 is presented in a matched paper. It involved psychometric validation of the pilot PROQOL instrument on 826 HIV patients across 8 countries.
The results section describes the themes that emerged from participant interviews that formed the basis of the PROQOL questionnaire.
Sociodemographic and clinical characteristics of the PWLHA who were interviewed are shown in Table 1. On average, participants were about 40 years of age, living with a partner and child, had been diagnosed for 7 years, and had contracted HIV heterosexually. However, participant characteristics varied considerably by country. For example, US participants were mostly older men diagnosed more than a decade ago who had contracted HIV through homosexual activity. They more often lived alone than with a partner or child. However, Senegalese participants were mostly women who had contracted HIV through heterosexual activity and were at an advanced stage of the disease. Very few lived alone or with a partner only, most residing with their children.
Thematic analysis of interview data revealed 7 HRQL issues that were important to patients but not covered by any single existing HIV HRQL instrument: fear of infecting others, concerns for the future, satisfaction with care, self-esteem problems, sleep problems, work disruption, and treatment issues. These issues were subsumed within 11 broader underlying themes that emerged to represent a conceptual model of HIV-specific HRQL: General health perception, social relationships, emotions, energy/fatigue, sleep, cognitive functioning, physical and daily activity, coping, future, symptoms, and treatment. The salience of particular themes and rationale for including particular items in the source questionnaire is described below.
General Health Perception
Across all countries, patients frequently and spontaneously made reference to their general health, and considered even slight improvements or declines as important to their sense of quality of life.
Interviewees described the quality of relationships with friends, family, and partners as very important to HRQL. This included receiving support from significant others, feeling comfortable, and socially accepted despite illness and treatment side effects. It included issues concerning actual and perceived stigma, involved fear of HIV transmission to others, rejection, level of isolation, changes in love relations, restrictions in sexual activity, and relationship difficulties arising from disclosure. Internationally, family discrimination was of particular concern in Senegal and India. Broader issues of stigma were more commonly raised by participants living in Senegal, China, Cambodia, and Thailand. Disclosure of HIV status seemed easier and more frequent in high-income countries (ie, Australia, France, and United States).
PLWHA described a myriad of negative self-perceptions and emotions affecting HRQL. Feelings of shame, guilt, inferiority, inadequacy, or embarrassment were common. Sadness, anxiety, irritability, and stress were pervasive. Participants in Cambodia were especially vocal about such issues.
Reports of low mental and physical energy were tied to quality of life perceptions. Participants described high levels of fatigue, needing to rest more often, the need for greater mental and physical effort to organize and perform normal daily routines, and to maintain pre-illness social activity levels.
During the interviews, many participants indicated difficulties with falling asleep, as well as reduced sleep time. Pain was a frequently cited cause, as was rumination and obsessive thinking about HIV seropositivity and its life consequences.
A number of participants related periods of time in which they experienced memory problems, and had difficulty remaining attentive and concentrating on activities, difficulties that had developed since contracting HIV or starting treatment.
Physical and Daily Activities
HRQL was commonly affected by limitations in physical activity. The interviewees raised issues concerning reduced capacity for home duties. Walking short distances, carrying light objects, and climbing stairs were described as taxing. Most participants were significantly limited in their capacity to perform strenuous activity. Walking limitations were of particular concern in low-income countries, where walking is the primary mode of traveling. However, travel difficulties linked to driving (an activity combining coordination, concentration, and physical ability) were often mentioned by Australian and US participants.
A number of participants described themselves as coping well with the effects of HIV. In broad terms, those who were able to integrate their HIV status and treatment routines into normal daily life described themselves as coping well. Religious or spiritual beliefs were also considered beneficial to coping. Participants in Australia, Brazil, and the United States were most likely to endorse these perspectives. The value of religious beliefs was also held by interviewees from Thailand and Senegal.
PLWHA often commented that they feared for their future because of HIV infection. Difficulties in planning for the future were commonly expressed. Concerns regularly centered on fears that their health condition would worsen, that normal life would become progressively disrupted, that opportunistic infections or other infections like the flu might occur, and that death may result. Australian, Brazilian, Indian, and Thai participants expressed fewer fears for the future when compared with other countries.
HIV symptoms and treatment-related side effects turned out to be a key quality of life issue. Pain, digestive problems, skin problems, weight loss or gain, changes in appetite, difficulties after prescribed diets, and changes in body shape or image were issues regularly raised. Treatment side effects received most attention in high-income countries, whereas in Cambodia, China, India, Thailand, and Senegal, modern treatment was described as a “savior” and lack of medical supply was feared.
PLWHA frequently expressed concern with the impact of HIV treatment regime on QoL. The necessity of taking medication at regular daily intervals was experienced as stressful. Specific concerns included fears of forgetting to take medication at the correct time, leaving medication behind when traveling away from home, and hiding consumption from public view to prevent questioning. Size and number of pills were also cited as problems by some.
In addition to these 11 themes, 3 independent issues were highlighted. First, satisfaction with health care services and relations with health care providers were common points of discussion. Second, financial and employment problems caused by HIV were regularly raised, including medical expenses, changes in work hours, and change or loss of job. The latter was of particular concern in Senegal, where 90% of participants were unemployed, which contrasted with interviewees in developed countries where the majority were actively working. Finally, interviews in several countries revealed worries about raising children, and in particular, conceiving children, because of HIV status. This issue was especially salient in Cambodia and Thailand where family lineage was considered very important.
Items by Theme
Agreement on the content of the 70-item draft questionnaires in English and French saw 67 items distributed across the 11 aforementioned themes. Three additional items were retained to cover the 3 independent issues of satisfaction with health care, financial difficulties, and parenthood. Example items for the 11 themes are shown in Table 2.
This study describes the first stages in the simultaneous development of PROQOL-HIV across 9 countries. PROQOL-HIV is a novel multidimensional HIV-specific HRQL instrument that strives to be sensitive to sociocultural context, disease stage, and treatment in the HAART era. During the development process, important new HRQL issues were uncovered from the culturally diverse experiences of PLWHA in previously underrepresented populations. To date, no other HIV HRQL measure has been submitted to such rigorous development across so many different countries.
In the very early stages of development, it became clear that important cultural differences in the HRQL concept needed to be considered. A single global question, “how is your quality of life,” was understood by PLWHA in some countries (United States, Australia, France, Brazil) but misinterpreted as “the standard of living” or not comprehended at all in others (Thailand, Cambodia, China, Senegal, India). Moreover, in these latter countries, it was difficult for some participants to distinguish between the direct impact of HIV on quality of life issues, and the impact of other concerns including chronic unemployment, health care access, and resource scarcity. Participants across all countries found it difficult to distinguish between HIV symptoms and treatment symptoms. Interviewers helped participants to separate the issues. This confirmed that HIV HRQL cannot be measured by a single question. Working simultaneously across the 9 countries from the early stages we succeeded in dealing with these issues while maintaining just one common questionnaire. These observations emphasize the importance of assessing HRQL as a multidimensional construct, clearly defining HRQL in culturally relevant terms, and delineating between HIV-related issues and other quality of life issues—particularly when developing a crossculturally applicable HIV HRQL instrument.
Through a process of literature review, consultation with experts and interviews with patients internationally, we identified 11 HRQL themes that broadly encapsulated the HIV experience. Subsumed within these themes were 7 issues important to HIV patients yet absent from any single existing instrument. The widely used MOS-HIV, for instance, does not incorporate HRQL disturbances resulting from pain-interrupted sleep, treatment issues such as the size and quantity of pills, or treatment effects such as body fat redistribution.6 WHOQOL-HIV does not address these issues either, focusing instead on medical dependence. WHOQOL-HIV further includes domains that are rather generic and not specifically linked to HIV, such as home environment, physical environment, and physical safety and security.19 Incorporation of the newly identified issues into HRQL measurement via PROQOL is clearly a step forward in accounting for the long-term experience of living with HIV in the HAART era.
Content validity in the theme-related items was also successfully maintained while incorporating cultural nuances. This was accomplished by finding common threads in issues across cultures while remaining vigilant to differences. In a small number of cases, this required wording items differently for particular countries. For instance, it was important to emphasize religious coping over spiritual coping in the case of Thai items, because in Thai Buddhist cosmology, spirits are viewed as potentially intrusive agents requiring regular appeasement. Likewise, extramarital sex and homosexual activity are not recognized in Muslim Senegal. Consequently, relational items explicitly referred to “husbands” or “wives” rather than “partners” in Senegalese-French and Wolof versions of PROQOL. Thus, although some items were necessarily formulated differently, they retained the same conceptual significance and related to the same themes across languages. With very few exceptions therefore, the PROQOL translations are conceptually identical.
The items also account for the experiences of patients at different disease stages, with varying access to treatment, within developed and developing countries. Indeed, a major strength in the development of PROQOL is the inclusion of countries such as Cambodia, India, and China, where there is greater heterogeneity in access to care, where treatment often occurs at a later disease stage,20 and where some patients (eg, untreated with low CD4) remain underrepresented in high resource countries like Australia, France, and the United States.
Some limitations need be acknowledged. A degree of subjectivity was inevitable during the translation process, interview analyses, and the process of identification and creation of themes. Other translators and researchers might have proposed slightly different nomenclature or issue categorization. Some questionnaire items were formulated slightly differently because of cultural sensitivities. PROQOL-HIV is a disease-specific instrument. It is not designed for use in general population studies or to compare the HRQL of populations with different pathologies.
In conclusion, PROQOL-HIV shows much promise as an HRQL instrument reflective of the themes that dominate the experience of HIV patients living in the HAART era, and important issues that are not measured by existing instruments. It has been developed simultaneously across 9 countries, following rigorous international standards. Psychometric validation of the PROQOL pilot instrument and determination of the dimensional structure are the subject of another article.
This study was made possible by the following people: Simon Mallal, Noel Hyland, Institute for Immunology & Infectious Diseases, Murdoch University and Royal Perth Hospital, Australia; Monica Barbosa de Souza, Jorge Marcio, Ana Lucia Weinstein, Sandra Barros Telles, Projeto Praça Onze, Hospital Escola São Francisco de Assis, Universidade Federal do Rio de Janeiro, Brazil; Ban Boroath, Kim Bopiseth, Vutha Nhao, Sary Sar, Rotha Heng, Service de Maladies Infecieuses, Hôpital Calmette, Phnom Penh, Cambodia; Zhu Zangping, 3rd People's Hospital, Hengyang, China; Jiqiang Fang, Jia Weidong, Cai Weiping, He Haolan, Gong lan, 8th People's Hospital, Guanzhou, China; Liu Jun, Center for Disease Control and Prevention, Hengyang, China; Yanbo Zhang, Jinfang Zhao, Tian'e Luo, Jian Sha, Leilei Pei, Hongmei Yu, Li Xiyan, Lu Yao, HY Tsui, Center for Disease Control and Prevention, Yuncheng, China; Joseph Tak Fai Lau, Xiaou Su, School of Public Health and Primary Care, Faculty of Medicine, The Chinese University of Hong Kong, China; Christophe Lalanne, Catherine Acquadro, AP-HP, Department of Clinical Research, PRO Unit, Saint Louis Hospital, Paris, France; Corinne Taeron, Miguel Ange Garzo, Hughes Fischer, Patient Associations, France; Patrizia Carrieri, INSERM UMR 912 SE4S Marseille, France; Agnés Levy, Antoine Béclère Hospital, Paris, France; Social Science Working Group, ANRS, Paris, France; Anne Persoz, Alioune Blondin-Diop, Patricia Assal, Ségeral Olivier, Sylvie Cheneau, Valérie Morin, Marie Thérèse Rannou, Marie-Stéphane Nguessan, Martine Mole, Sophie Ismael, Juliette Bailon, Sandrine Pottez, Yann Quertainmont, Cécile Goujard, Jean-François Delfraissy, Bicetre Hospital, Paris, France; Pascale Leclerc, CHU Grenoble, France; Christine & David Ellis Translation, Lyon, France; Veronica Noseda, Vincent Douris, Marc Dixneuf, Paola de Carli, Sidaction, Paris, France; ESTHER, Paris, France; Sharvari Apte, Vinod Bhalerao, Pallavi Nimbalkar, Girish Rahane, Madhuri Khaire, Somnath Borude, National AIDS Research Institute, Pune, India; Salif Sow, Bintou Dia, Adji Fatou, Abdoulaye Thiam, Moussa Diallo, Bineta Diallo, Cheikh Diop, Regional Research and Training Centre for HIV and Associated Diseases, Department of Infectious Diseases, Fann Teaching Hospital, Dakar, Senegal; Virat Klinbuayaem, Suwalai Chalermpantmetagul, Surush Sununta, Institut de Recherche pour le Développement, Research Unit 174 (IRD/Faculty of Associated Medical Sciences, Chiang Mai University/Harvard, School of Public Health), Chiang Mai, Thailand; Institut d'Etudes Démographiques, Paris, France; Kimberly Saulsberry, Robert Murphy, Center for Global Health, Division of Infectious Diseases, Northwestern University, Chicago, IL,; TransPerfect, New York.