The questionnaire was administered to all 74 eligible MOHs and 64 responded with a response rate of 86.5%. Out of the respondents, majority (62.5%) were males. This was in line with the proportion of MOHs who are males (i.e., 60.8%). Nearly half of the participants (i.e., 48.4%) were between 31 and 40 years of age. The most frequent total service period (of 26.6% of participants) was between 6 and 10 years. It was similar to 25% of MOH population having service period between 6–10 years.
The outputs of the correlation analysis are shown in Table 1. The correlation coefficients of ‘remuneration’, ‘recognition’, ‘work schedule’, and ‘responsibility’ showed positive relationships and were statistically significant. The strongest positive significant association was shown between’recognition and retention’ with a strength of 0.547 at a 0.001 significant level. (r = 0.547, p < 0.001). A positive and a statistically significant correlation exists between ‘responsibility and retention’ (r = 0.487, p < 0.001), ‘remuneration and retention’ (r = 0.439, p < 0.001), ‘work schedule and retention’ (r = 0.422, p = 0.001).
The VIF values were below 5 and tolerance statistics were above 0.4, which suggest the non-existence of multicollinearity within the data (Table 2).
The regression analysis (Table 3) revealed that the regression coefficients of ‘recognition’, ‘work schedule’, ‘remuneration’ and ‘responsibility’ were all statistically significant at 0.001 level. Thus, all four hypotheses were positive in direction and were statistically significant.
The regression analysis showed the highest magnitude of 1.039 representing the ‘work schedule’. It indicates that a positive relationship between work schedule (i.e., no duties public holidays, no on-call and night duties) and retention in the preventive health sector. This indicates that the medical officers are satisfied with the current work schedule in the preventive health sector.
The second highest magnitude of 0.564 was for ‘recognition’ (i.e., recognized by the patients, community, family members and others in the medical profession). This indicates that there is a positive relationship between recognition and retention.
A magnitude of 0.535 was seen for remuneration (i.e., take home salary, overtime payments, holiday payments and travelling allowance). This indicates a positive relationship between the level of remuneration and retention.
Responsibility showed a magnitude of 0.440 (i.e., responsibilities at the clinic, field and community work and having to work when off duty if need arises). This reveals a positive relationship between responsibility and retention.