1. Framework [1] demand for health services Perceptions of health care need to prevent, treat and rehabilitate a situation that has broken the health of people. This need to respond to physical factors (ie demand for health by accident, pregnancy, disease) or to life cycle factors life. Total
defendant
health care level of health that a person wishes to acquire.
premises of the model to estimate demand Given a perception of need, people can access the health system according to their budget constraint and take a fee for access to the availability of infrastructure and use it as the degree of need and degree of risk aversion.
Demand for health services has some elements not shared by the demand for most goods and services, which usually makes its modeling and estimation, these are:
a) Demand health has a significant random component.
b) The types of services demanded by household members are qualitatively different.
c) Demand for services behaves symmetrically rate movements, demonstrating discontinuities in behavioral function. Is explained by the conditionality of being sick, be assured, by income levels, the type of disease, the quality of services and habits.
d) Demand for services is associated with a sequential decision by the people.
• In a first step it is decided whether or not seeking health care, which is conditional on a perceived symptoms of illness, an accident, or require preventive controls.
• In a second stage, they decide to choose a type of health service provider (eg, MOH, ESSALUD or private).
to here refers to the demand for access.
• Once you access the service, determine the expenditure required for health care or intensity of use. This corresponds to the concept of demand is to use a "proxy" of the volume of health services used by an individual or a household, based on the aggregation of the actual cost of the basket of health services (outpatient, inpatient, ancillary tests and medicines).
demand model of health services
is an econometric model developed for estimating the demand is made by a sequential estimation equations:
· Model of illness perceptions
· Model of health system access
The equations have been modeled considering the methodology Heckman two-stage and estimated in the context of a probabilistic model.
2. The illness perception model
The decision of individuals to self-reported sick is selective.
A differentiated pattern of self-reported illness that varies by level of risk of the individual (related to living conditions), its position in the life cycle (age) and according to socio-economic conditions of the individual (such as sex, education level, employment status, insurance status) and home (such as gender and educational level of the individual who makes decisions , poverty status, the domain or geographic area to which it belongs, among other factors).
The patient self-reported decision is modeled using a qualitative dependent variable model (probit).
each individual is assumed to perform a cost-benefit analysis to make a decision.
The value of net profit declared for each individual patient is an unobserved variable (Z *). However, it is possible to observe the decision taken by each individual (it is observed whether or not self-report as patients). This decision is captured by a variable z that takes the value 1 when the individual is declared sick and 0 otherwise.
(Z *) = Ie + b d e
I + Z = 1 if Z *> 0
Z = 0 if Z * <0 ie=" vector" he=" vector" d="vector" b="vector" e="error">
3. Variables Model 1: perception of disease
Dependent Disease, 1 = sick, 0 = other
Definition: illness is chronic or sick in the last 4 weeks
Budgetary Poverty Level: 1 = extremely poor, 2 = not extreme poor, 3 = poor
dependency ratio
Active Population / No. household members
Definition: EAP is the economically active population, 1 = Dependent, 2 = Independent, 3 =
unemployed, 4 = PENALTY (not economically active population) Working
last 7 days, not working, unemployed seeking work, and
those without work are not looking because they know they will not find.
Insurance Definition: The probability of being insured net of the effects of PEA variables, sex and age
assurance model with these three variables, calculates the probability given these three factors
, given the difference of error
access costs Definition: weighted average rate of the basket of health services with day hospitalization
Hosp. (Day consultation fee). View number of days. Consultation
auxiliary tests (total analysis)
individual attributes
Sex: 0 = male, 1 = female
Age: 1 = between 7-25, 2 = between 26-60, 3 = under 6 and over 60
Attributes decision unit
Sex of head of household: 0 = male, 1 = female
Year Study of the head of Externalities
Risk of life: 1 = low risk, 2 = medium risk, 3 = high risk
Average of four indicators: a) time sources and average water supply,
b) type of toilet c) overcrowding (family members per room)
d) poverty level (as proxy for household food conditions)
Interactions
Rate Risk * insurance * insurance
extreme poverty rate * Price * Price * domain
nopobres
4. Model 2 variables: health service use
Dependent health service use, 1 = sick, 0 = other
Definition: patients who used health services, inpatient or outpatient
Budgetary
poverty level, 1 = extremely poor, 2 = poor not extreme, 3 = poor
dependency ratio per capita expenditure
:
Active Population / No. household members
Definition: EAP is the economically active population, 1 = Dependent, 2 = Independent, 3 =
unemployed, 4 = PENALTY (population economically not active) Working
last 7 days, not working, seeking work Unemployed and
those without work are not looking because they know they will not find.
Insurance Definition: The probability of being insured net of the effects of PEA variables, sex and age
assurance model with these three variables, calculate the probability given
these three factors, given the difference
error
access costs Definition: weighted average rate of the basket of health services with days of hospitalization
Hosp. (Day consultation fee). View number of days.
Consultation
ancillary tests (total analysis)
average per capita expenditure on health / ill
individual attributes
Sex: 0 = male, 1 = female
Age: 1 = 7 to -25, 2 = between 26-60, 3 = under 6 and over 60
Severity of illness
chronic discomfort
Sick Sick Injured
not chronic
Attributes
decision unit
Sex of head of household: 0 = male, 1 = female
Year of study of the household head
Externalities
Risk of life: 1 = low risk, 2 = medium risk, 3 = high risk
Average of four indicators: a) sources and the average time water,
b) type of toilet c) overcrowding (family members per room)
d) poverty level (as proxy for household food conditions)
Interactions
Rate Risk * insurance * insurance
extreme poverty rate * Price * Price * domain
nopobres
5. Estimation of pent-up demand for health services
The repressed demand estimation is performed using a sensitivity analysis on the model of block 1. The following table shows the probability of self-reported illness is higher in the non-poor when there are barriers, and change when you remove the economic barriers (pretending that the tariff is zero and that all are safe). It is noted that when you remove the economic barriers increases the self-reported disease in the extremely poor reaching levels similar to the non-poor (Figure 1). This suggests that the pattern of reporting of disease is affected by economic factors.
result, pent-up demand (the difference in the probability of self-reported sick with barriers less likely without barriers) is 10% in the group of extreme poor.
Chart 1
Estimate Potential Demand for health services
The potential demand for health services is derived from the sum of effective demand (which was declared in the survey) claim more covert (which did not use the service despite having declared to be ill due to lack of money) and more pent-up demand (the probability of declaring sick if remove economic barriers and uninsured).
