Short Communication

Urban–suburban differences in GP requests for lumbosacral spine radiographs in a primary healthcare centre in Malta


Glorianne Pullicino1 MSc, MRCGP, MMCFD, Assistant Lecturer *

Philip Sciortino2 MSc, MRCGP, FMCFD, Senior Lecturer and Head of Department

Sean Francalanza3 MD, General Practitioner Trainee

Paul Sciortino4 Undergraduate student, Medical Student

Richard Pullicino5 MD, MSc, MRCP, FRCR, Fellow in Interventional Neuroradiology


1, 2, 3, 4 Department of Family Medicine, University of Malta, Tal-Qroqq, Msida, Malta

5 Radiology, The Walton Centre, Liverpool, United Kingdom

ACCEPTED: 20 September 2017

early abstract:

Introduction: due to demographic changes, growing demands, technological developments and rising health care costs, analysis of resources in rural and urban primary care clinics is crucial. However, data on primary care provision in rural and suburban areas is lacking. Moreover, health inequities in small island communities tend to be reduced by social homogeneity and an almost indiscernible urban-rural difference. The aim of the study was to examine the urban-suburban differences in the indications for lumbosacral spine
radiographs in a public primary health care centre in Malta.
Methods: a list of all patients who underwent lumbosacral spine radiography in a public primary healthcare centre between January and June 2014 was obtained. The indications for lumbosacral spine radiographs were compared against the evidence-based indications posited by the America College of Radiology, the American Society of Spine Radiology, the Society for Pediatric Radiology, and the Society of Skeletal Radiology in 2014. Differences between suburban and urban areas were analysed
using chi-squared test. Direct logistic regression was used to
estimate the influences of different patients' characteristics and imaging indications in urban and suburban areas.
Results: the logistic regression model predicting the likelihood of different factors occurring with suburban patients as opposed to those residing in urban areas contained four independent variables (private/public sector, examination findings, osteoporosis, infection). The full model containing all predictors was statistically significant, 2 (chi-squared) (4, N = 1112) = 26.57, p = <.001, indicating that the model was able to distinguish between patients residing in rural and urban areas. All 4 of the independent variables made a unique statistically significant contribution to the model. The model as a whole explained between 2.4% (Cox and Snell R square) and 3.6% (Nagelkerke R squared) of the variance in suburban/urban areas, and correctly classified 78.5% of cases. All 4 of the independent variables made a unique statistically significant contribution to the model. GP requests for patients residing in suburban areas were more likely to be submitted from the private sector whereas urban GPs tended to include more examination findings. Requests by GPs for lumbosacral spine radiographs due to osteoporosis and infection tended to be more prevalent for urban patients.
Conclusions: such findings provide information for policy makers to improve equity in health care and resource allocations within the settings of urbanity and rurality.