The regression analysis showed that change in the number of unmet needs (determinant variable) was significantly associated with QoL at T1, a decrease in the number of unmet needs predicts higher QoL T1 scores.
Discussions were addressing the topics of: general aspects of quality in primary health care; possibilities to receive/provide PHC services based on both parties needs; determinant factors of accessibility to PHC services; patient centeredness.
First, our ability to adjust for important non-need determinants of utilization was limited.
Purging the effects of need factors on utilization of bias resulting from non-need determinants is critical to estimating need from utilization data[ 5, 18].
For Approach 2, we included multiple independent variables, in addition to age and sex, as indicators of need and non-need determinants of health care utilization [Additional file 1].
The first step, using individual-level data representative of the target population, estimates a model of the form: (1) where y i is the utilization for individual i, A ij is a vector of age-sex dummies, X ik is a vector of additional needs indicators, Z il is a vector of non-need determinants of utilization, and the R im are dummy variables for regions.
Ideally, the utilization model should include both need and non-need determinant of utilization in order to get unbiased estimates of the coefficients of the need variables [ 5, 9, 13].
The framework consists of three elements, including: assessment of needs or determinants of decisions, provision of decision support, and evaluation of the decision making process and outcomes of decisions made [ 15].
It also enables a closer focus on socioeconomic disadvantage and health need as determinants of utilization.
Studies such as intervention and prognostic studies can be performed in cohort and case-control designs, but need more determinants and confounding variables in their analyses.
