(1) Categories of health workers.
The first step in supply analysis is to decide whom to count. The most attention should be given to the categories which are costly to society (total numbers x average earnings). Therefore doctors, nurses, dentists, and technicians should be included. Midwives, herbalists, and similar categories should be counted only in countries where they are numerically important. Most drivers, hospital maids, and clerks should not be studied in detail, as they are part of a larger general labor pool. In this step it is highly important to have a clear definition of the categories of health workers (i.e. who should be included under the term “doctor”, or “chiropractors”, etc). For each class of health worker, three broad groups – professional, assistant, and aid level – should be adequate for general health personnel planning. Usually a definition based on years of training and income level is the most useful in determining these groups (however it is difficult to consider it as useful and adequate in Armenia at this time).

(2) Counts of workers.
Sources of current data on current supply, as well as the accuracy of the data differ in different countries. The sources may include

a) training institutions (a count of the past graduates, corrected for migration, deaths, and retirement from the profession is used)
b) licensing institution (gives the maximum numbers of legal practitioners in the country)
c) professional registries
d) government payrolls
e) the registry of private physicians
f) a special registration for narcotic use (may be useful in some countries lacking general registration)
g) the list of health workers paying professional taxes
h) rosters of professional societies
i) census data (but information can be five years, out-of-date, and not verified)
j) pharmaceutical companies lists of private practitioners (may be incomplete, companies may refuse to share info)
k) special survey which is the last and most accurate method of determining the current supply of health workers in the country. In the case of absence of a good source of the information, it could be the starting point for an effective registration system, however it is time-consuming and expensive

(3) Characteristics of health workers.
Several factors (such as age, sex, educational background, income, type of the practice/number of work hours, specialization, productivity – i.e., number of patients seen by the average practitioner per unit of time, and geographic location) should be considered when making predictions for change in supply.