C Robertson, P McConville, P Lefevre
Royal Cornhill Hospital, Aberdeen ; Andrew Duncan Clinic, Royal Edinburgh Hospital, Morningside Terrace, Edinburgh
Correspondence to: Dr Carol Robertson, Block A, Royal Cornhill Hospital, Aberdeen AB25 2ZH
E mail: Carol.Robertson@gpct.Grampian.scot.nhs.uk
SMJ 2005 50(2): 75-76
Abstract
Background and aims: Patients attending Accident and Emergency (A&E) on a frequent basis consume a large amount of NHS resources and often frustrate health workers employed in the service. This audit aimed to identify the personal and psychiatric characteristics of these patients and highlight areas where intervention may be helpful. Method and results: Patients presenting to A&E more than 20 times in four years were comparison to non-frequent attenders using case register information. The study demonstrates that frequent attenders are well known to psychiatric services, but are not engaged in ongoing contact. This is not due to a difference in attendance rate between frequent attenders and controls but may represent reluctance for follow up either on the part of the patient or the psychiatrist. Conclusions: These patients have a combination of physical, psychiatric and social pathologies and the services to address these needs does not seem to currently exist but should be considered.
Key words: Frequent attenders, A&E, psychiatry
Introduction
Frequent attenders at Accident and Emergency (A&E) account for a disproportionately high percentage of the total attendances and have an excess mortality over other patients.1,2 In the primary care setting, frequent attenders have higher rates of psychiatric symptoms.3 We investigated psychiatric morbidity among frequent attenders at A&E in comparison to non-frequent attenders.
Subjects methods and results
Aberdeen Royal Infirmary Accident and Emergency Department (A&E) covers the City of Aberdeen and surrounding region (population approximately 500,000).5 Index subjects were identified as those who had twenty or more attendances in the four year study period, 1September 1995 to 31 August 1999.
For each subject, two controls were identified as the closest age and sex matched individual attending A&E on the same day as the index subject’s first attendance within the study period. Medical records staff extracted from the Aberdeen Royal Infirmary Patient Administration System (ARIPAS) and the North East Scottish Psychiatric Case Register (NESPCR) details pertaining to each A&E attendance and each episode of psychiatric contact. Staff extracted data blind to index or control status. Statistical analysis is by Chi-squared tests for categorical data and by t-test for interval data.
A total of 114 frequent attenders and 228 controls were identified (75% male, mean age 40.85). These individuals had attended a total of 4442 times in the period, an average of 10 attendances per person per year. The highest number of attendance’s was 191.
Characteristics of their A&E attendances and psychiatric contacts are shown in Table I. Frequent attenders have a higher incidence of a “no fixed abode” residence and unemployment. Alcohol was identified significantly more often in the A&E attendance’s of frequent attenders as was deliberate self harm.
Frequent attenders are well known to psychiatric services, over 90% have had at least one episode of contact as compared with only 28.9% of non-frequent attenders and show 11.4 times the number of contacts as controls. Diagnoses of substance misuse, deliberate self harm and medical disorder are more common in frequent attenders, episodes of affective and neurotic disorder less common. There was no difference between the diagnoses of personality disorder. Fewer frequent attenders were offered follow up psychiatric appointments and their duration of contact was shorter than for controls, a difference that persisted once single assessments were excluded.
Comments
This study cannot comment on the age and sex profile of frequent attenders as our control group was matched for these factors. Our definition of frequent attenders, in the absence of any established criteria, is arbitrary and our choice of study period opportunistic. Data such as employment status, alcohol use and deliberate self harm were poorly recorded on the A&E database and may result in an underestimate of the number of attendances related to these factors. NESPCR fails to record a diagnosis in 28.7% of individuals and 47.7% of all psychiatric contacts, and does not distinguish between the omission of a diagnosis and a firm ‘diagnosis’ of no psychiatric illness.4
Nevertheless this study demonstrates that frequent Frequent attenders at A&E Robertson, McConville, Lefevre attenders are well known to psychiatric services, but are not engaged in ongoing contact. This is not due to a difference in attendance rate between frequent attenders and controls but may represent reluctance for follow up either on the part of the patient or the psychiatrist. We demonstrate that, for this patient group psychiatric services appear to have little to offer beyond assessment.
ACKNOWLEDGEMENTS: The authors would like to thank Dr Rainer Goldbeck, Dr John Eagles, Mr J Fergusson, Dr Elspeth Pitt and the Records Staff at Royal Cornhill Hospital and Aberdeen Royal Infirmary for their assistance with this project.
REFERENCES
1 Murphy AW, Leonard C, Plunkett PK, Brazier H, Lynam F & Bury G. (1999) Characteristics of attenders and their attendances at an urban accident and emergency department over a one year period. Journal of Emergency Medicine; 16:425-427.
2 Hansagi H, Allebeck P, Edhag O & Magnusson G. (1990) Frequency of emergency department attendances as a predictor of mortality: nine-year follow-up of population-based cohort. Journal of Public Health Medicine; 12 (1): 39-44.
3 Karlsson H, Lehtinen V & Joukamaa M. (1995) Are frequent attenders of primary health care distressed? Scandinavian Journal of Primary Health Care, 13: 32-38.
4 McConville P & Walker, NP. (2000) The reliability of case register diagnoses: a birth cohort analysis. Social Psychiatry and Psychiatric Epidemiology, 35: 121-127.
5 1991 Census