Assessment of staffing needs for physicians and nurses at Upazila health complexes in Bangladesh using WHO workload indicators of staffing need (WISN) method

Objective This study aimed to assess the current workload and staffing need of physicians and nurses for delivering optimum healthcare services at the Upazila Health Complexes (UpHCs) in Bangladesh. Design Mixed-methods, combining qualitative (eg, document reviews, key informant interviews, in-depth interviews, observations) and quantitative methods (time-motion survey). Setting Study was conducted in 24 health facilities of Bangladesh. However, UpHCs being the nucleus of primary healthcare in Bangladesh, this manuscript limits itself to reporting the findings from the providers at four UpHCs under this project. Participants 18 physicians and 51 nurses, males and females. Primary outcome measures Workload components were defined based on inputs from five experts, refined by nine service providers. Using WHO Workload Indicator of Staffing Need (WISN) software, standard workload, category allowance factor, individual allowance factor, total required number of staff, WISN difference and WISN ratio were calculated. Results Physicians have very high (WISN ratio 0.43) and nurse high (WISN ratio 0.69) workload pressure. 50% of nurses’ time are occupied with support activities, instead of nursing care. There are different workloads among the same staff category in different health facilities. If only the vacant posts are filled, the workload is reduced. In fact, sanctioned number of physicians and nurses is more than actual need. Conclusions It is evident that high workload pressures prevail for physicians and nurses at the UpHCs. This reveals high demand for these health workforces in the respective subdistricts. WISN method can aid the policy-makers in optimising utilisation of existing human resources. Therefore, the government should adopt flexible health workforce planning and recruitment policy to manage the patient load and disease burden. WISN should, thus, be incorporated as a planning tool for health managers. There should be a regular review of health workforce management decisions, and these should be amended based on periodic reviews.

The next step was data validation and set activity standards. Primary data validation was done 244 through phone calls made to the services providers, and health facility statisticians. Secondly, these 245 were shared with the SC and TT members. Finally, interviews were conducted with the EWG 246 members to finalize the activity standards (Table 1). We presented the time-motion findings to the EWG members and requested them to be as realistic 251 as possible in suggesting activity standards. We also requested them to account for the variability 252 of patient load due to factors such as seasonality, timing of day, facility catchment population, etc. 253 Finally, through multiple meetings, debates and deliberations, the activity standard was finalized, 254 taking into account the information from the IDIs and KIIs, and inputs from the EWG members 255 ( Workload components were defined through the inputs from the key informants; activity standards 284 were also set through the interviews with the EWG members. An activity standard is the time 285 necessary for a well-trained, skilled and motivated worker to perform an activity to professional 286 standards in the local circumstances 17 . Both service standards (for health service activities), 287 category allowance standards (for support activities), and individual allowance standards (for 288 additional activities) were determined in the same way.

524
Availability of data and materials 525 As our study is not involved in developing new software, or new database, therefore, the datasets 526 generated and analyzed during this study (which was developed using WISN software) are not 527 publicly available. But if necessary, we will share our database for future research purposes.     Determining priority cadre(s) and health facility types(s) Priority cadres and facilities (from both communities and facilities) were determined after discussing with the Steering Committee members.
Estimating available working time Was decided based on the findings from the document reviews, key informant interviews and in-depth interviews.
Defining workload components Was determined based on key informant interviews with the experts, supplemented by observations and in-depth interviews with active service providers.
Setting activity standards These data were collected through both in-depth interviews and a quantitative technique, time motion study.
Establishing standard workloads A standard workload is the amount of work within a health service workload component that one health worker can do in a year. This was calculated using WISN software.
Calculating allowance factors This is to document the additional and support activities performed by a health staff.
This was also calculated using the WISN software.
Determining staff requirement based on WISN Through secondary data extraction from health facility records, annual service statistics were collected in order to determine the staff requirement Analyzing and interpreting WISN results  Comparing the difference between current and required staffing levels, we identified the health facilities that are relatively understaffed or overstaffed  Using the WISN ratio, we assessed the work pressure that health workers experience        UpHCs. This reveals high demand for these health workforces in the respective sub-districts.

44
WISN method can aid the policy-makers in optimizing utilization of existing human resources.

45
Therefore, the government should adopt flexible health workforce planning and recruitment policy  Bangladesh's health workforce scenario is characterized by "shortage, inappropriate skill mix and 85 inequitable distribution" 8, 9 . Equitable access to skilled and motivated health worker in a functional    The research continued from July to November 2017 and was carried out in two pre-selected  availability of service statistics. An 'hourglass' approach was adopted for defining the workload 220 components based on the ESP (Figure 4). Tools were pre-tested in a UpHC near Dhaka, before 221 applying for actual data collection. Qualitative tools were also pre-tested through mock IDIs and 222 KIIs. The pre-testing exercise was followed by the training of the Field Supervisors (FSs) and    We presented the time-motion findings to the EWG members and requested them to be as realistic 249 as possible in suggesting activity standards. We also requested them to account for the variability 250 of patient load due to factors such as seasonality, timing of day, facility catchment population, etc.

251
Finally, through multiple meetings, debates and deliberations, the activity standard was finalized, 252 taking into account the information from the IDIs and KIIs, and inputs from the EWG members 253 ( In order to ensure the quality of data, the Principal Expert (lead author of this article), Co-experts 262 (two co-experts -one was leading data collection and the other was leading data quality check and   additional activities) were determined in the same way.

Availability of data and materials
KIIs. The pre-testing exercise was followed by the training of the Field Supervisors (FSs) and   We presented the time-motion findings to the EWG members and requested them to be as realistic 249 as possible in suggesting activity standards. We also requested them to account for the variability 250 of patient load due to factors such as seasonality, timing of day, facility catchment population, etc.

251
Finally, through multiple meetings, debates and deliberations, the activity standard was finalized, 252 taking into account the information from the IDIs and KIIs, and inputs from the EWG members 253 ( In order to ensure the quality of data, the Principal Expert (lead author of this article), Co-experts 262 (two co-experts -one was leading data collection and the other was leading data quality check and  additional activities) were determined in the same way.