PATTERN OF REFERRALS OF HEAD INJURY TO THE UNIVERSITY COLLEGE HOSPITAL, IBADAN

---

A.O. Adeleye

Department of Surgery, College of Medicine, University of Ibadan, and University College Hospital, UCH, Ibadan, Nigeria

Abstract

Background: Although there are now many neurosurgical units in Nigeria, cases of head injury (HI) continue to present in the University College Hospital, UCH, Ibadan, from hospitals all over the country.

Objective: This report aims to highlight the contemporary patterns of referrals of cases of head injury for neurosurgery in Ibadan.

Methods: The study emanated from an analysis of registry of cases of head injury managed in our practice covering a 7-year period at the UCH Ibadan. The clinical/ trauma-demography of cases of head injury referred were analyzed quantitatively and then compared with head injury cases that presented primarily to UCH Ibadan during the same period. An alpha value of <0.05 was considered significant for associations for pattern of referrals.

Results: Of the 1034 cases of head injuries in the registry, 23.6% presented primarily to UCH, Ibadan; 423 (40.1%) were from outside Ibadan/Oyo state. Most of the cases were earlier seen in other health facilities in four out of Nigeria’s six geopolitical zones including other teaching hospitals with practicing neurosurgeons. The reasons for these inter-hospital, inter-state referrals included absence of neurosurgical expertise (67%) or lack of other logistics like neuroimaging, bed space and intensive care unit services. Head Injury was caused by road accidents in more than 85% of the cases. The patients referred inter-state had more severe injuries, more delayed attainment of critical milestones in their trauma care, and higher frequency of neurosurgical operative interventions. They also had worse in-hospital outcome and longer length of stay.

Conclusions: The University College Hospital, Ibadan continues to play a pivotal role in the management of head injury in Nigeria. The results support an urgent need for stakeholders in the health sector to provide all that is required to uphold the status of the Hospital as a center of excellence in neurosurgery and neurosciences.

Keywords: Head injury, Referrals, Neurosurgery, UCH, Ibadan

Introduction

In the developing countries, also known as low-middle income countries (LMIC), head injury (HI) constitutes a sizeable proportion of the clinical caseloads of most neurosurgical units.^1-3 The optimal care of many cases of HI, especially the severe ones, is time bound. The time was put at 4 hours for traumatic acute subdural haematoma in a landmark publication some four decades ago.⁴ In spite of the establishment of new neurosurgical units in many parts of Nigeria in recent years, referrals for the neurosurgical care of HI still present daily in the University College Hospital (UCH), Ibadan from many regions of the country. Many of these arrive late, and in sub-optimal clinical conditions.

This is a descriptive cross-sectional survey of a HI registry with the aim of establishing the pattern of referrals of HI for neurosurgery in Ibadan. This, it is hoped, would provide insights into more purposeful allocation of health care resources for the management of HI in Nigeria.

MATERIALS AND METHODS
This was a prospective study of consecutive HI patients managed in a neurosurgical practice at the University College Hospital, Ibadan between August 2009 and June 2016. The database of these cases of HI was examined for the referral patterns. Firstly, the cases of head injury presenting primarily and secondarily to the University College Hospital, Ibadan were sorted. Cases of HI that presented directly at the UCH, Ibadan were termed primary presentation. Those that had been seen in another health facility before presenting in UCH were the secondary ones. The sources of referrals of the latter were noted including types of health facilities (private, general hospitals or other tertiary facilities) from which they were referred for neurosurgical care, and the state of origin in Nigeria where the trauma occurred. The number of health facilities (one, two and more) earlier visited before arrival in UCH, Ibadan, as well as whether there was any in-house neurosurgical service in any of these was also noted. The reason for the inter-hospital transfers from these other centers was also noted from the referral letters when available.

Clinical/Statistical analysis
This was followed by an analysis of the demographic, clinical and trauma characteristics of the referred cases which were analyzed and compared with those of the cases presenting primarily in the UCH, Ibadan. This was with a view to noting any associations and or differences in the clinical-demographic trauma profiles between the two patient groups. Some of the parameters analysed in this respect were many of the known clinical determinants of outcome from HI: the presence of loss of consciousness (LOC), hypothermia (systemic temperature <35.00C) hypotension/hypertension, tachycardia/bradycardia, high fever (systemic temperature >38.50C), anaemia (packed cell volume less than 30%), pupillary anomalies/asymmetry of reaction (anisocoria), and the severity of the HI using the Glasgow coma scale (GCS).
The Injury Severity Scores (ISS) providing a measure of severity of the overall injuries for patients with multiorgan trauma was also computed for each case. The ISS is scored from 0-75, higher figure denoting increasing severity of trauma.⁵. Finally, the patients’ in hospital outcome at discharge or death was quantified using the dichotomized Glasgow outcome scale (GOS). Good outcome were cases in ‘normal status’ or ‘moderate deficits’ while poor outcome were those with ‘severe deficits’, ‘vegetative states’ and ‘death’.⁶

Data management
The clinical records of the cases had been captured consecutively in clinical summary forms, and were transferred to an electronic spreadsheet, using the SPSS version 21 (The SPSS Inc, Il, USA). It was thus a primary database. The data were analyzed with the same software and presented here in descriptive statistics including frequencies and proportions, means (standard deviation, SD), 95% confidence interval, CI, median (range). Categorical variables were explored for associations with the chi-squared test. Associations between parametric continuous variables were explored with the 2-tailed student-t test, and those between non-parametric variables with the Mann- Whitney-U test. An alpha value of <0.05 was deemed significant for associations.