This study was conducted at Kisumu County, Kenya in facilities recognized by Ophthalmic Division. Kisumu County was chosen as it is the pioneer county to roll Universal Health Coverage hence need of optometrists to aid in achieving the Universal Health Coverage objectives. The main economic activity is fishing, a clear indication of long exposure to sunlight which is a risk factor for cataract. Kisumu County has a population of 1.2 million (Census, 2019). The study was conducted in the following facilities: Sabatia eye hospital, Jaramogi Oginga Odinga Teaching and Referral, Monitor Optical, Optic Ophthalmology Centre, Vision Eye centre, Port Florence Clinic and City Centre Optics.
This was a prospective cross-sectional study. The design was more appropriate as it provide accurate account of the characteristics of a particular individual for the purpose of discovering new meaning and describing what exists.
Condition for recruitment was based on registration by Optometrists Association of Kenya and working in facilities recognized by the Ophthalmic Division of Kenya. The optometrists included in the study were aged between 25-36 years. The sample was derived from a target population of 149 registered optometrists with the Optometrists Association of Kenya. The participants were recruited from June 2019 to August 2019. Using standard normal deviate of 1.96; a conservative proportion of 0.5 in the target population (given no previous knowledge, skills and practice reference in Kenya) were estimated to be well conversant with cataract and a level of accuracy required fixed at 0.05; a sample size of 49 participants (using a base population of 149) was determined using formula:. Optometrists who were not registered were excluded from the study.
Participant’s recruitment was done through an invitation letter sent to the optometrists explaining what the study was all about and what the significance of the study was. Three reminders were sent to the participants through their email informing them on when the study will be conducted.
Simple random sampling was conducted to recruit participants from the respective clinics (n= 7) and demographics. From a sampling frame of n=149 optometrists, a sample size of 49 optometrists was derived. The researcher listed the target population and assigned consecutive numbers from 1to 149. An online random number calculator was used to generate the numbers to be included in the study. A total of 49 random numbers were selected from the sample frame which constituted the sample size. The research instrument was pretested among 5 optometrists. A pilot study with a sample of a tenth of the total sample is appropriate for a pilot study (Mugenda & Mugenda, 2008). Therefore a tenth of the total sample was 5. The participants included in the pilot study were not included in the actual study because they would have influenced the outcome the results. The researcher tested reliability using Cronbach’s alpha (0.874, 0.929 and 0.926 for knowledge, skills and practice questionnaire respectively) and validity using a Pearson correlation coefficient (0.000<0.05, N=5).
After potential participants were identified, they consented and a copy of the signed consent was given to them. The response rate was 100% and this was enhanced by maintaining a constant contact between participants and the researcher at all time during the study period.
Data collection instrument
The aim of this study was to assess the optometrists knowledge, skills and practice on cataract as it is one of the main causes of blindness. Self administered questionnaires were used it had potential in reaching out to large number of respondents within a short time and it gave the respondents adequate time to respond to the items. Seven research assistants who were optometrists were recruited to administer the questionnaires to the respondents. The research assistants were trained on research ethics and how the process was to run. The questionnaires were given to only optometrists practicing in the seven recognized Ophthalmic Division facilities in Kisumu County, Kenya.
The questionnaires consisted of closed ended questions worded in present tense aiming to explore the participant’s knowledge, skills and practice on cataract. Information collected included: socio demographic characteristics, knowledge of various aspects of cataract, their skills on cataract and practice on cataract management. The study had 16 structured questions broadly around respondent’s demographic characteristics (3 questions), knowledge on cataract (4 questions), skills on cataract (7 questions) and practice on cataract (2 questions). The questions had two to three responses of aware, neutral and not aware on a three point Likert scale. One open ended question was on the reason for referral of cataract cases. The questionnaires took 20 minutes to complete. A composite scale ranging from 20 to 100 was adopted with knowledge, skills and practice level categorized as: low (score of 20-40), medium (score of 41-79) and high/good (score of 80-100). The participants were to identify the type of cataract based on the seven photographs shown to them. The questionnaires were administered when participants were free and collected back each and every evening. The questionnaires were given to the participants and they were given an option to seek clarification in case of any difficulty from the lead researcher
Frequency distributions of all socio demographic characteristics and the proportion of participants who identified all items related to skills, knowledge and practice were calculated. A chi-square was used to calculate the association between demographic characteristic and knowledge, skills and practice. We conducted multiple logistic regression analysis to compare the knowledge, skills and practice with demographic characteristics. We calculated odds ratios (ORs) and 95% confidence intervals (CIs). Thematic analysis was conducted to analyze the question on reason for referral. There was no missing data as all data obtained were stored manually and in the computer. Statistical Package for Social Sciences version 17 software was used to analyze the data. Values of p < 0.05 were considered statistically significant.