2.1 Sampling
In this study, we adopted a random cluster sampling method. Among six sample provinces, two provinces were selected for this survey in the eastern, central and western regions (Eastern: Zhejiang and Fujian; central: Anhui and Henan; western: Yunnan and Shaanxi). In each province, we chose five prefecture-level cities at random and in each prefecture-level city, we randomly selected two counties. Five THCs were randomly chosen from each county. In total, 300 THCs from six provinces were surveyed. On average, each THC has 40 on-duty staff, 34 health care specialists, 16 doctors and 5 certified (assistant) doctors, with an average of 37,270 outpatient and emergency visits plus 1,344 inpatients per year. An average of 20 administrative villages and 16 village clinics are under the governance of each THC. Each village clinic is staffed by 1.3 health care professionals.
2.2 Data collection
Using the NEPHS guidelines for 2011, the research team designed questionnaires and survey instructions (see the questionnaire on job responsibilities of village doctors in the supplementary material). Survey questions covered basic information of the sampled THCs and statistics of the total workload assigned according to the 12 types of services and the share of workload undertaken by village doctors in village clinics. The share of workload undertaken by village doctors in village clinics was estimated and reported separately by each service-related department in THCs according to the actual workload carried out by village doctors. The survey tool was pre-tested in two THCs of Beijing and further revised. Before conducting the survey, the research team confirmed the list of sample provinces, cities and counties. To guarantee a high response rate, the Primary Health Department of the National Health and Family Planning Commission issued a notice regarding the survey. Provincial health and family planning administrative departments in the sample provinces organized sample counties and selected sample THCs, as required, checking whether all answers on the questionnaires had been completed and reviewing the data for accuracy. After receiving the questionnaires, the research team rechecked everything and carried out logic checks. During the survey, the research team provided advice by telephone and kept records.
2.3 Model
The equivalent value (EV) method has been used to estimate the cost of the NEPHS program and to calculate community health-staffing requirements [28-32]. In the present study, we used the EV method to build a model for measuring the workload of NEPHS provided by village doctors, according to the following steps: 1) determine the standard service protocol of all types of NEPHS; 2) determine the workload and EV of each NEPHS compared with a standard clinic visit; 3) calculate the village doctors’ workload in the NEPHS program.
Step 1: Determining the standard service protocol
In 2016, the NEPHS program had 12 types of public health services. These included establishing health records for residents; management of patients with chronic non-communicable diseases; physical examination for major diseases in children, women and older people; health education; vaccination services for vaccine-preventable diseases and prevention and control of major infectious diseases. All 12 types of NEPHS are included in the 2011 NEPHS guidelines [11].
Step 2: Determining the workload and EV of each NEPHS
To determine workload (person-time), we used a multistage iterative feedback and revision process [31-33]. Participants (n=60) from the six sampled provinces were invited to attend a series of five meetings according to their expertise with NEPHS. Participants included THC managers (n=12), public health workers (n=12), family physicians (n=12), nurses (n=12) and village doctors (n=12). During the meetings, participants discussed the amount of person-time required for each NEPHS, according to the 2011 NEPHS guidelines. Participants also suggested modifications to the workload indicators. Socioeconomic education levels among villagers as well as population density in western provinces are lower than those in the central and eastern provinces. Participants generally believed that these differences affect the delivery of NEPHS, thus affecting their workload [31-33]. For example, for rural residents in western China, a clinic visit or home visit may require more time from village doctors. Additionally, NEPHS may be provided using mobile medical facilities to assure service accessibility, which will also have an impact on the NEPHS workload. The gap in the workload of NEPHS between eastern and central China was relatively small, so these were combined into one region, and two sets of specific workload were created for each NEPHS.
To test the workload of each NEPHS, four research assistants were trained to observe and record the person-time for each type of service in 12 THCs randomly chosen from the six sampled provinces. In terms of services that could not be recorded during direct observation, face-to face interviews with public health workers were conducted to determine their workload. The person-time for each NEPHS was rechecked and modified on the basis of direct observation and interviews.
To ensure that different public health services could be compared directly, a “standard clinic visit” was introduced as a benchmark to gauge the EV for NEPHS [28-33]. A standard clinic visit referred to a family physician consulting with one patient for 15 minutes [34] and the EV of a standard clinic visit was defined as 1. The EV of each NEPHS was then determined based on the person-time compared with a standard clinic visit. The workload and EV of each NEPHS in different areas was defined separately and is shown in Table 1.
Step 3: Calculate village doctors’ workload in the NEPHS program
Based on the EV of each NEPHS, the workload undertaken by village doctors under the NEPHS program was calculated using the following process:
Share of the workload undertaken by village doctors under the NEPHS program = EVs of services performed by village doctors under the NEPHS program (X) / total EVs of services included in the NEPHS program provided by sampled THCs (Y) × 100%
Y= ∑ EV of each public health service item (A) × volume of each public health service item (B)
X= ∑ EV of each public health service item (A) × volume of each public health service item (B) × village doctors’ share of workload for each public health service item (C)
A: sourced from the EV of each NEPHS in the Table 1;
B: sourced from surveys on the total workload assigned according to the 12 types of service (see section 2.2, Data Collection);
C: sourced from surveys regarding the share of workload undertaken by village doctors in village clinics. This was estimated and reported separately by each service-related department of THCs according to the actual workload conducted by village doctors (see the Data Collection section).