Isaac Macharia

Swahili Translation and Validation of the Patient Health Questionnaire-9 Depression Scale in the Kenyan Head and Neck Cancer Patient Population

Background: Depression is an important predictor of post therapy quality of life (QOL) in head and neck (H&N) cancer patients. In addition, depression rates may vary among cultures. Objective: As part of a larger cross cultural study on post therapy QOL differences in H&N cancer patients, the goal of this project was to translate a well-validated English language depression scale into Swahili, and then validate this scale in Kenyan H&N cancer patients. Methods, Settings and Subjects: In Part 1 of the study, we translated the Patient Health Questionnaire-9 (PHQ-9) into Swahili, adhering to established International Quality of Life Association (IQOLA) guidelines. In Part 2, we psychometrically validated the newly translated scale using a prospective study of 48 patients at the Kenyatta National Hospital ENT clinic in Nairobi, Kenya. Results: The Swahili PHQ-9 had good test retest reliability (Intraclass correlation coefficient, 0.71) and internal consistency (Cronbachs alpha=0.80). It also had good construct validity, as scores correlated strongly with TNM stage (Chi square=123, p < 0.05), and with the composite and global scores of an H&N cancer specific QOL scale (UW-QOL, r = −0.87, p < 0.05). Conclusion: The Swahili version of the PHQ-9 is a reliable scale in Kenyan H&N cancer patients, and is a valuable tool in screening for and monitoring of depression as a function of QOL in this population.

A prevalence study of ear problems in school children in Kiambu district, Kenya, May 1992

Information on the prevalence of hearing impairment and related ear pathologies in children in sub-Saharan Africa is scarce. A pilot study for a clinical trial of simple treatments for chronic suppurative otitis media (CSOM) in school children in Kiambu district, Kenya, provided information on the prevalence of hearing impairment and ear pathologies. Five-thousand-three-hundred-sixty-eight children from 57 randomly chosen primary schools in Kiambu district were examined. Simple otoscopy was performed by clinical officers with specialty training in ENT, and hering testing was performed by trained nurses, using a hand held field audiometer. Microbiological specimens were obtained from those children with CSOM. Five-point-six percent of the children had a hearing impairment of > 30 dB HL in one or both ears, with 2.2% having bilateral hearing impairment. Two-point-four percent had at least one perforated tympanic membrane, and 1.1% had CSOM. Eight-point-six percent of the children had wax obstructing the tympanic membrane. There is evidence of a relationship between hearing impairment and both CSOM and wax obstructing the tympanic membrane. The most common organisms found were Pseudomonas spp. (34%), Proteus spp. (34%) and Eschericia coli (19%). These results are comparable with other studies in Africa and indicate a considerable burden of ear disease in Kiambu district, Kenya.

Randomised controlled trial of treatment of chronic suppurative otitis media in Kenyan schoolchildren

BACKGROUND The outcomes of treatment of chronic suppurative otitis media (CSOM) are disappointing and uncertain, especially in developing countries. Because CSOM is the commonest cause of hearing impairment in children in these countries, an effective method of management that can be implemented on a wide scale is needed. We report a randomised, controlled trial of treatment of CSOM among children in Kenya; unaffected schoolchildren were taught to administer the interventions. METHODS We enrolled 524 children with CSOM, aged 5-15 years, from 145 primary schools in Kiambu district of Kenya. The schools were randomly assigned treatments in clusters of five in a ratio of two to dry mopping alone (201 children), two to dry mopping with topical and systemic antibiotics and topical steroids (221 children), and one to no specific treatment (102 children). Schools were matched on factors thought to be related to their socioeconomic status. The primary outcome measures were resolution of otorrhoea and healing of tympanic membranes on otoscopy by 8, 12, and 16 weeks after induction. Absence of perforation was confirmed by tympanometry, and hearing levels were assessed by audiometry. 29 children were withdrawn from the trial because they took non-trial antibiotics. There was no evidence of differences in timing of withdrawals between the groups. FINDINGS By the 16-week follow-up visit, otorrhoea had resolved in a weighted mean proportion of 51% (95% CI 42-59) of children who received dry mopping with antibiotics, compared with 22% (14-31) of those who received dry mopping alone and 22% (9-35) of controls. Similar differences were recorded by the 8-week and 12-week visits. The weighted mean proportions of children with healing of the tympanic membranes by 16 weeks were 15% (10-21) in the dry-mopping plus antibiotics group, 13% (5-20) in the dry-mopping alone group, and 13% (3-23) in the control group. The proportion with resolution in the dry-mopping alone group did not differ significantly from that in the control group at any time. Hearing thresholds were significantly better for children with no otorrhoea at 16 weeks than for those who had otorrhoea, and were also significantly better for those whose ears had healed than for those with otorrhoea at all times. INTERPRETATION Our finding that dry mopping plus topical and systemic antibiotics is superior to dry mopping alone contrasts with that of the only previous community-based trial in a developing country, though it accords with findings of most other trials in developed countries. The potential role of antibiotics needs further investigation. Further, similar trials are needed to identify the most cost-effective and appropriate treatment regimen for CSOM in children in developing countries.

Evaluation of the use of a questionnaire to detect hearing loss in Kenyan pre-school children

In developing countries, there is a lack of trained personnel and testing equipment to facilitate the early detection of hearing impairment in children. A questionnaire offers a low cost option and the value of this for detecting hearing impairment in pre-school children was determined in several districts in Kenya. The questionnaire was completed by either teachers, parents/carers or community nurses. The children were subsequently tested using pure tone audiometry and visual examination of the ear by ENT Clinical Officers, who were not given prior access to the results of the questionnaire. A total of 757 (88%) questionnaires were completed. Of the 735 children, who could be tested using pure tone audiometry, four were found to have a unilateral hearing impairment and one was detected by the questionnaire. A total of 13 children had a bilateral hearing impairment >40 dB HL. All were detected using the questionnaire. There were eight males and five females with ages ranging from 4.2 to 6.9 years, mean age 5.7 years and median age 5.8 years. Eight had a sensorineural hearing impairment and two a mixed hearing impairment. Three of the children with a sensorineural hearing loss had a family history of hearing impairment. No question detected all children with a hearing impairment and some questions were more discerning than others. There was 100% sensitivity for the questionnaire when a hearing loss of >40 dB was considered, but specificity was lower at 75%. Negative predictive value was 100%, but the positive predictive value was low, 6.75%. It was concluded that a questionnaire of this nature could be usefully applied at Primary Health Care level for detecting hearing impairment at the pre-school stage. There would be need for services available for diagnosis, treatment and habilitation before a screening programme was introduced.

Topical quinolone vs. antiseptic for treating chronic suppurative otitis media: a randomized controlled trial.

Objective  To compare a topical quinolone antibiotic (ciprofloxacin) with a cheaper topical antiseptic (boric acid) for treating chronic suppurative otitis media in children.

GDP matters: cost effectiveness of cochlear implantation and deaf education in sub-Saharan Africa

Hypothesis Cochlear implantation and deaf education are cost effective in Sub-Saharan Africa. Background Cost effectiveness of pediatric cochlear implantation has been well established in developed countries but is unknown in low resource settings, where access to the technology has traditionally been limited. With incidence of severe-to-profound congenital sensorineural hearing loss 5 to 6 times higher in low/middle-income countries than the United States and Europe, developing cost-effective management strategies in these settings is critical. Methods Costs were obtained from experts in Nigeria, South Africa, Kenya, Rwanda, Uganda, and Malawi using known costs and published data, with estimation when necessary. A disability adjusted life years (DALY) model was applied using 3% discounting and 10-year length of analysis. Sensitivity analysis was performed to evaluate the effect of device cost, professional salaries, annual number of implants, and probability of device failure. Cost effectiveness was determined using the WHO standard of cost-effectiveness ratio/gross domestic product per capita (CER/GDP) less than 3. Results Cochlear implantation was cost effective in South Africa and Nigeria, with CER/GDP of 1.03 and 2.05, respectively. Deaf education was cost effective in all countries investigated, with CER/GDP ranging from 0.55 to 1.56. The most influential factor in the sensitivity analysis was device cost, with the cost-effective threshold reached in all countries using discounted device costs that varied directly with GDP. Conclusion Cochlear implantation and deaf education are equally cost effective in lower-middle and upper-middle income economies of Nigeria and South Africa. Device cost may have greater impact in the emerging economies of Kenya, Uganda, Rwanda, and Malawi.

Pattern of congenital neck masses in a Kenyan paediatric population

BACKGROUND The distribution of congenital neck masses varies between countries and is important in diagnosis and treatment modalities. Data from Africa is scarce, and altogether absent from Kenya. OBJECTIVE To describe the pattern of congenital masses in a Kenyan paediatric population. SET UP: Kenyatta National Hospital, Nairobi Kenya. STUDY DESIGN Prospective descriptive cross-sectional. METHODOLOGY Children aged 15 years and below with neck masses who presented to various surgical clinics and wards at the Kenyatta National Hospital between December 2006 and April 2007 were included in the study. They were examined for age, gender, type and location of the neck mass. Mode of investigation and final diagnosis were recorded. Data was analyzed by using statistical package for social sciences. Descriptive statistics were applied to determine means, frequencies and modes. Ninety-five percent confidence interval was used and p value of 0.05 taken as significant. Data are presented in form of tables. RESULTS Two hundred and thirty-five children (129 male) with neck masses were examined. Fifty-one (22%) of these masses were congenital. Thyroglossal duct cysts were the most common (29%) followed by cystic hygromas (21%) and branchial cleft cysts (20%). Fifty-one percent of the masses were present at birth. The midline was the most common location (31.4%) followed by anterior border of sternocleidomastoid (27.5%) and submandibular region (19.6%). Ultrasound was the commonest diagnostic investigation. CONCLUSION Congenital defects constitute an important differential diagnosis for paediatric neck masses in Kenya. Thyroglossal duct cysts, cystic hygromas and branchial cleft cysts are the most prevalent occurring most commonly in the midline and anterior border of sternocleidomastoid muscle. An understanding of the distribution of these masses improves diagnosis, preoperative decision making and their overall management.