Developing a measure of health-related quality of life for women with cervical dysplasia resulting from human papillomavirus infection

Participants

For open-ended and cognitive interviews, participants were recruited from the colposcopy clinic of a community hospital in an affluent suburb in the midwestern region of the USA. Participants provided informed consent before completing study interviews. Participants were women who attended a gynaecological oncology clinic, over 18 years of age, with evidence of cervical infection positive for HPV. A physician determined each participant to have one of the following: low-grade squamous intraepithelial lesions (LSIL) after adequate colposcopic examination ruling out high-grade disease or cancer, LSIL on cytology, atypical squamous cells of undetermined significance, or normal cytology and a risk factor indicating a high probability of persistent infection with high-risk HPV (e.g. age >30 years, still positive on repeat Pap smear testing). Participants with high-grade squamous intraepithelial lesions were excluded from participation. Funding for the project came from an industry project aimed at developing a topical ointment for women with low-grade HPV-related disease. Thus, we accessed a population with low-grade disease as we developed the preliminary questionnaire.

Interview procedures

In open-ended interviews, a masters level study coordinator, trained in qualitative methods and cognitive interviewing techniques, used a semistructured interview guide to query participants on how they defined quality of life and the most important physical and psychological symptoms they experienced. After completing open-ended interviews, 20 participants provided an importance rating on items from the draft pool of items from the literature review and items from the FACIT database. We averaged these importance ratings and revised or dropped items based on these responses: items with low importance ratings were dropped. Item content (overlap with existing items) and clinical relevance were also used to guide decision-making.

After a draft measure was developed, a new set of 10 participants, recruited from the same clinic, completed cognitive interviews to help ensure that items would be understood as intended. ¹³ The study coordinator queried participants on the language, comprehensibility and relevance of the items. Both the open-ended and cognitive interviews were digitally recorded and transcribed verbatim. These qualitative data were then content analysed such that the study director (DR) and coordinator (RK) read and memoed the transcripts, identified prominent themes based on well accepted qualitative data coding procedures and came to consensus on themes through a series of discussions. ¹⁴

Literature review

Nineteen articles in peer-reviewed journals were identified to help determine symptoms and concerns affecting people with HPV infection. Special effort was made to distinguish between disease-related and treatment-related problems. Final tabulation showed that 23 symptoms or concerns were documented (Table 1).

Clinician survey

The symptoms and concerns identified provided a guide to selecting items for the measure. Items were selected from the FACIT item library, and the set of items were then reviewed by three clinician-investigators who were proficient in the diagnosis and treatment of patients with HPV infection. The clinicians completed a self-administered survey in which they provided an importance rating on 111 potential items. This information was then used to reduce the number to a 68 item pool.

Open-ended interviews

Twenty participants with cervical dysplasia provided self-reported sociodemographic and clinical information, completed open-ended interviews and gave importance ratings on the 68 items in the pool (see Tables 2 and 3). Psychosocial concerns were prominent themes that emerged from the interviews, as participants had particular concerns around embarrassment, disclosure, social support and anxiety. From information gathered in these interviews and importance ratings, we narrowed the pool to the 33 items that became the draft measure. In order to address participants' concerns about the future, the item ‘I am hopeful about the future’ was included. In order to address participants' concerns that their infection would become cancerous, we included the item ‘I worry the infection will get worse,’ which was worded more generally to be applicable for people with all types of HPV infection.

Cognitive interviews

Sociodemographic information for the cognitive interview participants is provided in Table 2. Ten new participants first completed the 33 item draft measure, provided a severity rating on each item using Likert-type response categories (0 = not at all, 1 = a little bit, 2 = some-what, 3 = quite a bit, 4 = very much), and gave feedback on the understandability of the draft scale. Some participants did not have a partner or did not tell a family member about their infection. Therefore, we added two conditional questions to give participants the opportunity not to answer questions about partner and family if not applicable, which completed the final version of the scale (Table 4).

Overall, we began with 93 items that were reduced to 68 items based on expert review. Data from participants helped us to craft a more appropriate questionnaire. After incorporating participant importance ratings, the number items were dropped down to 33. Cognitive interview results suggested that we add questions to make the preliminary version of the scale contain 36 items.