Health worker incentives: survey report, Sierra Leone

January 2015

Authors: Sophie Witter, Maria Paola Bertone, Haja Wurie, Mohamed Samai, Joseph Edem-Hotah, Rogers Amara

Read the full text of this report here.

Background

The ReBUILD research programme focused on health system development in post-conflict countries. Our aim was to develop lessons for governments on how to make, recreate or sustain fair health systems. Sierra Leone was one of the focal countries of the research.

One component of the research looked at how to reach and maintain incentive environments for health workers to support access to affordable, appropriate and equitable health services. One of the research tools was a health worker survey, which is the focus of this report. The aim of the survey was to understand health workers working patterns, sources of remuneration and motivation, and how these had changed in recent years. Findings from this survey were  triangulated with other quantitative and qualitative research tools.

Methods

A structured questionnaire was administered to 312 health personnel from all of the key professional categories in four districts of Sierra Leone. It is estimated that the respondents covered 12% of the total staffing in these districts. The districts were chosen to be representative of the different regions and so they included urban/rural variations, remote/hard to reach areas, and varying measures of poverty/need. The study sites were: Western Area (Urban/Rural); Kenema District (Eastern Region); Bonthe District (Southern Region); and Koinadugu District (Northern Region). The field work was conducted in December 2012.

The questionnaire probed the following topics: respondents’ background and household characteristics; their employment, workload and working hours; training patterns; pay from different sources; motivation; views on how their working life has changed in recent years, factors motivating work in rural areas; and plans for the future. Reponses were analysed by profession, district, facility type and gender, using Stata.

Ethical approval was granted by the Sierra Leone Scientific and Ethics Committee and the Liverpool School of Tropical Medicine.

Study limitations included difficulty getting the full planned sample of health workers, especially for doctors and staff based in the Western Area. Some questions were also poorly understood and answered, leading to the need to exclude outliers. Locational data would have been a useful addition to the questionnaire. Finally, the tool relies on self-reporting, whose accuracy is hard to assess.

Findings

Respondent characteristics Respondent characteristics broadly followed the known national pattern, being dominated by SECHNs (100) and MCH aides (55), with fewer representatives of CHOs/CHAs (44), lab technicians (28), state registered nurses (25), pharmacists and pharmacy technicians (20), EDCU Assistants (16), environmental health officers (14), and doctors (11). Overall, 57% of respondents were female, but with wide variations according to professional categories. Most roles are male-dominated, with the exception of nurses/midwives and MCH aides.

The bulk of doctors and nurses were in Western Area, while other categories such as MCH Aides and CHOs/CHAs were more evenly distributed. This is a reflection of the current national HRH distribution.

Most respondents were in their forties and married with children (67%), though there was a significant difference by cadre, with MCH aides and EDCU assistants being generally younger (the majority of these cadres were in their thirties). Educational qualifications varied according to the different professions. The majority of the sample (62%, with no significant variation across cadres or genders) were working outside of their home district at the time of the survey.

Household characteristics

The mean number of people in the respondents’ household ranged from 6 in doctors’ households to 8 for lab technicians (overall range was 1-22). The mean number of dependents was higher, ranging from 6 to 11 across the professional groups (overall range 1-27). Analysis by district indicates that larger households and higher numbers of dependents are found outside the Western Area.

Significant differences were found in overall expenditure, with doctors spending over 3 million Leones per month on average, compared to 792,000 Leones for the EDCU Assistants. However, as a proportion of overall expenditure, the cost of food accounted for close to 50% of total expenditure for most groups. Differing patterns were found across the districts, with higher total expenditure per household in Western Area and Bonthe.

28% of respondents reported having saved money in the previous month, while 43% had received loans, with significant differences across the professions (doctors having the highest savings rates and lowest loan rates of all professions). Health workers outside Western Area were more likely to have borrowed in the previous months, but there were no significant differences by gender.

The most common asset owned by health workers’ households was a mobile phone (more than 98% overall), while the least common was a car (just over 10% overall). Across the professions, there were significant differences for electricity, TVs, videos, fridges, motorbikes and cars, with doctors and RNs having higher access to most assets, apart from motorbikes, which are more common for other groups. Less than one third of respondents had access to running water and around 39% had access to a fridge. Across the districts, significant differences were found, with higher access to electricity, TV, videos, fridges, running water and cars in Western Area. Conversely, motorbike ownership is highest in Koinadugu. Across the sexes, there are no significant differences except for motorbikes, which men were more likely to own.

Employment and workload

Respondent’s place of employment was closely linked to profession. All doctors were based in hospitals; most CHOs/CHAs were based in the CHCs; 80% of RNs were in the hospitals, while SECHNs and Environmental HOs were more evenly distributed across facility types; the bulk of MCH Aides and EDCU Assistants were based in the MCHPs; laboratory technicians were largely but not exclusively hospital-based, while pharmacists and their assistants were concentrated in the tertiary hospitals. 32% of respondents were based in regional, district and secondary hospitals, and just over 5% were in faith-based facilities.All were public servants (on the MoHS payroll), even those based in faith-based facilities.

On average, the HWs had spent just over 11 years working in the health sector, just under 11 years in the public sector and nearly 4.5 years in their current post. Between the professions, environmental health officers had worked the longest in the sector and in public service (around 19 years), as well as serving longest in their current posts, alongside pharmacists (both over seven years). Nearly 90% of all respondents had worked for the public sector in their previous post.

The average number of hours reported worked per week across all respondents is 54, with significant differences by cadre. CHOs/CHAs report the highest mean (65 hours per week). Some (e.g. the SECHNs) report very high maximums (168 hours per week), which may reflect the fact of being on call and living near facilities. Only 59 respondents (19%) stated that the hours worked per week had changed over the last 3 years. There were significant differences across professions, with MCH aides most likely to report an increase (25%). Of those who stated that their hours of worked had changed, 48 (81%) said their hours had increased, 9 (15%) said that they had decreased and 2 (3%) did not reply.

The average number of patients seen across all the respondents who answered was 116.6 (i.e. about 19 per day in a 6-day week), with SRNs seeing 90 patients per week and lab technicians seeing 190 patients (but clearly less intensive interactions in the latter case). 12 There is evidence of significantly higher workloads in Western Area, but increases were highest in Bonthe.

Training

85% reported receiving in-service training. Significant differences were found across profession for training at external universities. Doctors, RNs, CHO/CHAs and pharmacists were most likely to have received external training. Differences between districts were also found for local university training, with those in Western Area more likely to have received this. Some differences in access to training were noted between the genders (lower rates for women), though these were not significant.

Respondents were asked what they saw as the benefits of training. Greater knowledge, more confidence, and higher status came top (with 97%, 90% and 64% respectively). 19% reported that training increased private practice earnings. Across the cadres, significant differences are found for some attributes. For example, charging higher in private practice is cited by 45% of doctors, compared to 8% of registered nurses. Seeing more patients is also more frequently cited by doctors (73%).

Income from main work

Respondents were asked about their last month’s salary. The difference between doctors’ and other cadres’ salaries is striking, with doctors earning more than four times the salary of an RN, and RNs being paid almost three times the amount o EDCU assistants. Women are significantly less well paid in general, though this is not significant when broken into different cadres (i.e. reflects the employment mix), except for CHOs/CHAs, where women are paid significantly less. This may reflect different lengths of service.

Only 31% (96) of the respondents stated that their salary changed in the last 3 years. This was unexpected. However, a number of the health workers were volunteers before the salary uplift in 2010, and were only put on the MOHS payroll afterwards. They are therefore unable to report on changes related to the start of the Free Health Care Initiative itself.

Only 16% of respondents mentioned that they received a Remote Area Allowance (RAA). It seems that the majority of health workers are not aware that they are eligible and do not receive the payment. 71% of those who received the RAA did not do so regularly, and even those in receipt did not have a clear understanding of the nature of the RAA.

Salary supplementation was more popular prior to the FHCI for health workers attached to health programmes. Post-FHCI, this was minimised due to the salary uplift associated with the FHCI, so it is not surprising that less than 10% reported being in receipt of this. They were more common outside Western Area and came from donors and NGOs.

Only 4% of respondents stated that they received any payment from user fees in the last month. The low number of observations may be due to the fact that sharing user fees revenues is now not accepted by the Anti Corruption Commission (ACC) since the introduction of its Service Delivery Charter in September 2010, which is meant to encourage and support public institutions to become more transparent in their dealings with the public. All fees should now go to the consolidated fund of the facility. 5% reported other payments, such as gifts from patients.

At the time of the survey, 7 PBF payments should have been received by facilities and staff. A quarter of respondents in PHUs (which are eligible for payments) had received no PBF payments, while a third had received 3, with no significant difference by type of PHU but significant differences by cadre and district (with the highest payments being in Koinadugu). The pattern across districts is either linked to variable performance or problems of disbursement in certain districts.

Of the 310 respondents, 42% received a per diem/DSA in the last month. There was a significant difference by profession, with MCH aides most commonly receiving them in the past month (65%), followed by environmental health officers (57%), CHOs/CHAs (56%) and doctors (55%). Differences are also evident across the districts and facility types. Koinadugu and Bonthe report the highest frequency of per diems. The paying of per diems seems to be more concentrated in primary facilities. The maximum number of per diems received was 4 and the overall mean was 1.5 per diems per month for those reporting receipt. 34% of these were provided by the MoHS, 30% by NGOs and 27% by the UN. Significant differences are found across the cadres, with the largest amounts received by SRNs and doctors. Overall, 28% of respondents declared in-kind benefits. The most common was housing, reported by 14% overall, but most common for doctors and nurses. 5% report receiving in kind gifts from patients.

Overall totals show the difference in scale income from main job for doctors, as well as the dominance of the salary element. Apart from additional funds from per diems, other sources constitute a small proportion of total income from main job for public sector staff. Salary constitutes from 63% of main income for MCH aides to 92% for pharmacists. RAA ranges from 0% (for most cadres) to 8% for MCH Aides. PBF payments range from 0% (for doctors, RNs and pharmacists) to 16% for MCH Aides, user fees from 0% to 3%, per diems from 3% to 21%, top ups from 0% – 8%, and other sources from 0% to 2%.

Considering the total public pay per hour worked, doctors are paid 20,245 Leones, compared to 4,215 Leones for CHOs/CHAs. Pay per patient seen is even more differentiated, with a 30-fold difference between RNs and doctors, compared to a four-fold difference in pay per hour, 14 suggesting that RNs are seeing a higher volume of patients in their hours of work. Across the districts, there was no significant difference in payments per hour worked or per patient.

Private practice

Only 6% of respondents reported doing private practice. There was significant variation across the professions, with the highest rates reported for pharmacists (45%), followed by doctors (18%). As would be expected, there was also significant variation across the districts, with Western Area and Koinadugu reporting more private practice, which is almost entirely absent in Kenema and Bonthe. Men and those in tertiary facilities were significantly more likely to report private practice. Those who did report working in private practice did so for a relatively high number of hours per week, ranging from 9.5 for the doctors to 42 for one nurse. Private clinics and private pharmacies were the most common locations given for private practice, and were most commonly owned by someone else. Only 13 out of 19 disclosed their monthly income from private practice (no doctors disclosed).

Other income-generation

84 respondents (27%) stated that they have some income-generating activities (IGA) outside of the health sector, with differences which were not statistically significant across the professions. For the 53 who provided estimates of hours spent in other IGA, the range was from 3 hours per week as a mean for doctors to 12 per week for SECHNs. Trading and selling is the most common type of IGA reported by the group (65% of all IGA reported), followed by farming (29%). Lecturing is also an additional source of income for the doctors. Income generated from IGA per week shows a familiar gradient across the professions, though with the EDCU Assistants reporting relatively higher sums than, for example, nurses.

Total income

The difference in the total income from all sources between doctors and other professions is highly significant, ranging from 471,583 Leones as mean monthly income for EDCU assistants to 4,741,300 for doctors. There are no statistically significant differences in the income from the main employment by facility. However, there are significant differences between districts, with higher incomes in Koinadugu and Western Area, and by gender, with women earning less than men across cadres.

Looking at the composition of overall income, salary is dominant for all groups, especially for doctors, for whom it provides more than 80% of total income. For EHOs, the proportion is 55%. Pharmacists derive an important proportion of their income from private practice (around 18%). Per diems for workshops provide between 2% of income for pharmacists and 21% for Environmental Health Officers – 12% of income across all groups. For the staff in PHUs, the PBF component contributes about 10% for CHOs, SECHNs, EHOs and MCH Aides. RAA is only reported by CHOs, SECHNs, EHOs and MCH Aides and is limited in amount (about 6% of total income for these cadres).

Motivation to stay in post

When asked about what motivated respondents to stay in their job, opportunities to serve the community were most frequently citied (90%), followed by good relationships with colleagues (79%), the security of work (71%), opportunities for training (70%), social status (65%) and salaries (63%). When asked to rank them in order of importance, salary emerges as the most highly ranked, followed by opportunities for training and additional allowances/opportunities to serve the community.

By profession, when ranked, additional allowances came top for doctors and pharmacists, whereas salaries are first for most other groups (CHOs/CHAs, SECHNs, EHOs, MCH Aides, ECDU Assistants). RNs put opportunities for training as most important, while lab assistants put serving the community as their most highly ranked factor.

Perceived changes

Participants were asked how their life had changed in the last decade. Very few negative changes were reported – the main one being an increased workload, which was reported by 12.5% of respondents. The largest group (46%) reported personal benefits (such as greater confidence, peace of mind, self respect), followed by educational improvements (33%), and improvements to salary and living conditions (24%). 18% feel more able to care for their families. Relatively few (8%) think that working conditions have improved, while less than 2% report improvements in their ability to serve the community, and 6% report no change. Across the professions, significant differences in responses can be noted. Doctors, RNs, SECHNs, EHOs and pharmacists particularly highlight personal benefits, while CHOs/CHAs and MCH Aides are more aware of educational changes, and EDCU assistants reported improvements to salary and living conditions.

Willingness to work in rural areas

When asked an open question about what would motivate staff to serve in rural areas, the main themes which emerged were the following:

 Accommodation (housing, etc.)

 Financial incentives (salary increase, remote and/or risk allowance, incentives, etc.)  Support to family (school fees, scholarships for children, family and children facilities, etc.)

 Communication support (airtime, communication allowances)

 Transport support (transport allowance, mobility, motorbike, vehicle, fuel, etc.)

 Access to basic amenities (water, electricity, toilet, food, basic facilities)

 Training (more education and training for the health workers)

 Improved living conditions (improvement of social opportunities, social amenities, relationship with communities, etc.)

Promotions (linked to rural posting) 

 Investment in working conditions (improvement of working conditions, relationship with colleagues, support and supervision, more staff, equipment and drug availability, etc.)

 Provision of healthcare for health workers

Of these, financial incentives are cited the most as being important (80%), followed by better accommodation (64%), transport support (56%), improved working conditions (18%) and access to basic amenities (14%). All other factors are cited by less than 10% of the group.

Future plans

When asked an open question about plans for their future, 77% of respondents planned to develop their career or pursue their education, compared to 28% who were focussing on providing for their family, 13% who wanted to start a private business and 5% were approaching retirement.

Conclusions

To our knowledge, thiswas the first quantitative survey of its kind to be done in Sierra Leone, investigating how the main public sector health professionals live and work, how they earn their living (including from private practice and additional income-generating activities) and seeking their views on changes to their working lives, what motivates them and what would motivate them to work in rural areas. The survey provided policy-relevant evidence for the MoHS and its partners, particularly as they faced the challenge of rebuilding the sector post Ebola. The findings of the survey were triangulated with some of the other research tools which ReBUILD used, which include key informant interviews, a document review, routine HRH data analysis and life history interviews with health workers. In general, the findings cohered across the different tools. For example, in the life histories work, health workers reported that access to training opportunities are biased in favour of those in urban areas, which was supported by the survey findings that those in Western Area are more likely to have received university training and that doctors, RNs, CHOs and pharmacists were most likely to have received external training.

It is encouraging in general that, according to this survey, life for health workers had improved in general over the previous years (noting that the field work was done before the current Ebola crisis). This fittedwith documented changes in HRH policies since the introduction of the Free Health Care Initiative, which improved conditions for health staff. However, dangers were also highlighted by the increase in salaries, especially for higher cadres, which must be combined with continued efforts to improve the accountability and performance of staff. Comparisons with other studies in the region suggested that doctors in 17 particular are well paid in Sierra Leone and that the overall working hours and workload of staff remains reasonable.