Determinants of Healthcare Use Based on the Andersen Model: A Systematic Review of Longitudinal Studies

Abstract

The aim was to give an overview of longitudinal observational studies investigating the determinants of healthcare use explicitly using the Andersen model. To this end, three electronic databases (Medline, PsycINFO and CINAHL) were searched (and an additional hand search was performed). Longitudinal observational studies examining the determinants of healthcare use (outpatient physician services and hospital stays) based on the Andersen model were included, whereas disease-specific samples were excluded. Study quality was evaluated. The selection of studies, extraction of data and assessment of the studies were conducted by two reviewers. The following determinants of healthcare use were displayed based on the (extended) Andersen model: predisposing characteristics, enabling resources, need factors and psychosocial factors. In sum, n = 10 longitudinal studies have been included in our systematic review. The included studies particularly showed a longitudinal association between increased needs and higher healthcare use. Study quality was rather high. However, several studies did not conduct robustness checks or clarify the handling of missing data. In conclusion, this systematic review adds to our current understanding of the factors associated with healthcare use (mainly based on cross-sectional studies). It showed mixed evidence with regard to the association between predisposing characteristics, enabling resources and healthcare use longitudinally. In contrast, increased need factors (in particular, self-rated health and chronic conditions) were almost consistently associated with increased healthcare use. This knowledge may assist in managing healthcare use. Since most of the studies were conducted in North America or Europe, future longitudinal studies from other regions are urgently required.

Health Need Factors Are the Key Drivers of Hospitalization among the Elderly Living Alone: An Analysis of Longitudinal Data

“…Some of the variables of the Behavioral Model predicted overnight hospitalization for Australian elderly people living alone. However, as in previous research [ 10 , 14 ], the need factors were the key drivers. That is, across both sets of analyses (original data and the imputed data), illness severity with comorbidity (Time 1) and self-rated general health (Time 2) were the consistent, substantial predictors of overnight hospitalization.…”

“…For example, people on lower incomes, an enabling resource, may be less likely to have preventative screening [ 8 ], which may result in more severe health conditions (need) over time. There may also be broader contextual considerations, in that the health care systems of some countries reduce or remove the role of enabling issues (e.g., in Germany, income is not a driver of health care utilization [ 14 ]), presumably through the presence of relatively universal health care, widespread health insurance coverage [ 15 ], or veteran insurance programs [ 10 ].…”

“…There is a growing set of studies testing the Behavioral Model (e.g., see [ 14 ] for a review), and a novel focus here is on the social characteristics of those living alone. Social factors have been suggested to affect hospitalization rates for the elderly beyond specific medical conditions [ 8 ].…”

“…The relationships between social issues may also be overlapping to some degree. For example, social isolation may be associated with physical health and mental health [ 17 ], where social issues may be considered as enabling resources [ 14 ], but could also be impacting need factors. A possible explanation for these varying degrees of relationship between social issues and hospitalization is the breadth of social issues considered.…”

Longitudinal patterns and predictors of healthcare utilization among cancer patients on home-based palliative care in Singapore: a group-based multi-trajectory analysis

“…Andersen’s landmark model suggests health service use as a function of (1) predisposition to use services, (2) enabling or impeding factors, and (3) need for care. This model is widely used to investigate factors associated with healthcare utilization [ 36 , 37 ].…”