Study design, setting and participant recruitment
This was a cross-sectional study carried out between March and September 2020. It was a sub study of a larger study (hereby called parent study) entitled “Oral health quality of life and dental treatment needs among HIV + children and adolescents on ART attending selected HIV clinics in Kampala.” The study was carried out in Kampala, the capital city of Uganda. The participants were selected from four HIV clinics: Mulago Immune Suppressive Syndrome (ISS) clinic, Kawaala Health Centre IV, Kisenyi Health Centre IV, and Kiswa Health Centre III which acted as clusters. The sampling frame for this sub study was 400 adolescents aged 10–18 years as of March to September 2020 attending the four selected HIV clinics. Sample size was estimated using an online calculator at 95% confidence level, significance level 0.05, the calculation was based on the premise that estimated population proportion of HIV infected adolescents that have ever used dental care was 50% (25). The minimum sample size that was required for this study was 197 adolescents (25). A total of 246 adolescents were conveniently contacted according to availability of their telephone contacts and consent to participate in a telephone interview. Of these 44 participants had either wrong telephone numbers or their numbers were inaccessible. These 44 were excluded from the study and not included in the analysis. Another 48 participants were excluded from the study because they were not aware of their HIV serostatus. A total of 154 adolescents were included in the study and hence the participation rate was 62.6% (154/246).
Inclusion criteria for this sub-study included participants had to be part of the parent study and 10–18 years as of March to September 2020. They had to be HIV positive, aware of the HIV serostatus and attending HIV clinics Kawaala, Mulago, Kisenyi, and Kiswa, and whose telephone contacts were accessible and available at the time the study was conducted.
Interviews
This study was strictly based on telephone interviews due to the COVID-19 rules and guidelines (26). Each interview lasted for about 30 minutes. A structured questionnaire that guided the interview was used using the Open Data Kit form (27) that had been standardized for all participants. Two research assistants underwent a face-to-face training in using the Open Data Kit form and they were in addition trained on how to conduct telephone interviews in both English and Luganda. They were each given a description of the study objectives and study outcome for standardization of the tool. They were both given electronic tablets Samsung Galaxy SM- T285 S# R52J60MZ22V and R52JB215X8TA and trained how to use them. Both these devices were loaded with the standardized questionnaire in an ODK form and was used uniformly for all study participants.
The questionnaire included socio-demographic characteristics, factors affecting use of dental care services, personal dental health practices, general and HIV related concerns and attitudes and included the outcome variable “ever-visit- the dentist”.
Parents or caregivers were asked if the child was aware of their HIV serostatus at the start of interviews to exclude those not aware, as some parents withheld information about the serostatus and explained other reasons for frequent treatment. The questionnaire was translated to Luganda from English by a local translator. Later five independent people (the main project study investigator, 2 research assistants and 2 pilot study participants) reviewed the accuracy of the language translation. Back translation from Luganda back to English was carried out informally through analysis of questions if they meant the same in English. Therefore, the telephone interviews were conducted mostly in the locally spoken language, Luganda, and a few in English. Andersen’s behavioural model guided the selection of variables used to determine access and use of dental services among the selected group of participants (28)
Independent variables
The various variables were defined according to Andersen’s behavioral model into predisposing (socio-economic) factors, enabling factors, need related factors, oral hygiene, and personal health practices.
Among predisposing factors, gender was categorized into boys (0) and girls (1); age was grouped into 2 groups; 10–13 years (1) or 14–18 years (2). The level of education had 6 options, and these included: Primary 1–3 (1), Primary 4–7 (2), Senior 1–4 (3), Senior 5 or 6 (4), Vocational courses (5) and those not attending school (6), and these were categorized into 3; those that had no formal school (0) and those that were attending primary level of education (1) and those that were attending secondary level of education (2). Vocational courses were categorized under secondary level.
The socio-economic status was assessed using a wealth index that assessed ownership of household items as assessed by Filmer et al (29). The participants were asked if they possessed the items or not. A participant was recorded as possessing the item only if the item was functioning. The items included television, electricity, bicycle, water motor car, flush toilet, mobile phone, computer, radio, motorcycle, and refrigerator. Using principal component analysis (PCA), five quantiles were generated with 1 representing the poorest and 5 the richest quintile. These were further categorized into least poor (1) and poorest (0) quintiles. Home description had five options which included very good (1), good (2), bad (3), very bad (4) and I don't know (5). These were categorized into 2; good (1) and bad (2).
Under enabling factors, the questions concerning fear of going to the dentist had 3 options (1–3); I do not fear (1), I fear a little (2) and very fearful (3) and these were re-categorized into 2; “yes (1)” for both little fear and fear and “no (0)” for no fear. The questions concerning fear of HIV spread, avoidance of dental care due to HIV status, avoidance of dental care services due to cost, and failure to receive dental treatment because of other medical conditions had responses with 4 options (1–4) and these included yes, several times (1), yes, a few times (2), no, never (3) and I don’t know (4) and these were categorized into 2; “yes (1) or no (0)”, “yes” for both several times and a few times and “no” for “no, never”. All responses with “I don’t know” were added to the most frequent response.
The question regarding who is responsible for decision making in seeking dental care services had the following responses: parents (1), myself (2), my caregiver (3), my teacher (4) and I do not know (5). These were categorized into 2: parents/teachers/caregivers (0) and myself (1). Under need related factors, rating on general and oral health had five options. These included poor (1), fair (2) good (3), very good (4) and excellent (5). These were further categorized into three; poor (1), fair (2) or good (3). Satisfaction of oral health was assessed via four parameters, and these included very satisfied (1) satisfied (2), dissatisfied (3) and very dissatisfied (4) and these were categorized into two options either satisfied (1) or dissatisfied (0). “Satisfied (1)” category included responses of both very satisfied and satisfied while the “dissatisfied” category included responses of both dissatisfied and very dissatisfied. For the factors under personal oral hygiene and dental practices, the frequency of toothbrushing had 6 responses (1–6). These included the following: never (1), several times a month (2–3 times) (2), once a week (3), several times a week (2–6 times) (4), once a day (5) and 2 or more times a day (6). These were then grouped into two; those who rarely brush (0) and those that brush daily (1).
Items used for tooth brushing /tools for tooth brushing assessed included soap, salt, urine, local herbs, ash, toothpaste, and nothing. These items were renamed in Excel workbook 2003(*xlsx), and each item dichotomized into yes (1) for those who used that item for brushing or no (0) for those who did not use that item.
For those who had been to the dentist, responses concerning last visit to the dentist had 5 options (1–5) and these were the following: less than 6 months ago (1), 6–12 months ago (2), more than 1 year but less than 2 years ago (3), 2–5 years ago (4) and more than 5 years ago (5).These were sub-grouped into those that had visited the dentist less than a year ago (1), between 1–2 years ago (2) or more than 2 years ago (3). Reasons for the last dental visit included mandatory school check-ups or routine check-ups, emergency (tooth injury), emergency (toothache), having tooth (teeth) pulled, filling, root canal or others. were renamed and dichotomized into “yes (1)” for those whose responses were positive for a particular reason or “no (0)” for those with negative responses for a reason.
Duration at the dental facility had 5 options that were categorized into 3: less than 1 hour (1), 1 to 2 hours (2) or 3 and more hours (3). Average travel cost was assessed under 5 options (1–5). The options included the following: less than Ug. Sh. 4,000 (1), between Ug. Sh. 4100–10000 (2), no money spent (3), more than 10000 (4) and I do not know (5). These were categorized into; no money spent (0), less than Ug. Sh. 4,000 (1), between Ug. Sh. 4100–10000 (2) or more than 10000 (3). Attitude of the oral health care providers was also assessed 5 options (1–5) and these included very good (1), good (2), average (3), below average (4) and don't know (5) that were categorized into 3; good (1), average (2) or below average (3). “Very good (1) and good (2)” responses were put under the “good (1)” option.
NB: All “I don’t know” responses were added to the most frequent categories.