Outcome and appropriateness of referrals to emergency department

Tseng-Kwong Wong 黃增光, Yuk-Tsan Wun 溫煜讚, David V K Chao 周偉強

HK Pract 2005;27:286-293

Summary

Objective: Appropriate referrals to the heavily loaded Emergency Department are important for better utilization of resources. This study aims at assessing the appropriateness of referrals from a general outpatient clinic in terms of referral outcome at Emergency Department.

Design: A retrospective study.

Subjects: Referrals from a public outpatient clinic to a regional Emergency Department.

Main outcome measures: Management and route of discharge from the Emergency Department.

Results: 74.5% referrals were found to be appropriate using the defined criteria by outcome. The inappropriate referrals could be due to the referring doctors' inadequate diagnostic and/or management skills, insufficient time for some operative procedures, and the lack of facilities for observing/monitoring patients in the outpatient clinic.

Conclusion: Appropriateness of referrals to the Emergency Department could be improved by providing training to staff, optimizing consultation time and establishment of observation facilities at the clinic.

Keywords: Referral, family practice, emergency hospital service, outcome assessment.

摘要

目的： 適當的轉診對於有效運用已繁忙的急症資源至為重要。本文就政府門診轉診的洽當性加以評估。

設計： 回顧式研究。

研究對象： 某一政府門診向一所地區性醫院急症室的轉診情況。

主要測量內容： 急症室醫護人員對病人的診治以及離院安排。

結果： 74.5%的轉介按既定的標準評核為適當。不適當轉介原因包括轉介醫生缺乏足夠的診斷及／或治療技巧，沒有充裕的時間施行手術，沒有充足的設施來觀察這類病人。

結論： 通過培訓職員，給予充裕診症時間及在診所內建立觀察設施，或可以改善轉診至急症室洽當性。

詞彙： 轉介，家庭醫學，急症醫療服務，結果評核。

Introduction

Referrals are common and of variable rates in family practice.^1,2 In Hong Kong, studies showed that referral rates varied between 1.6% and 4.6% of consultations.³ The appropriateness of referrals from family practice varies among different studies from 90.4% in the UK to 75.4% in Canada.^4,5 In one local study 3% of all referrals were considered as "inappropriate" and 11% of "borderline" inappropriateness.³

Referrals to the Accident and Emergency Department (AED) deserve attention in Hong Kong as AED is heavily loaded.⁶ Inappropriate referrals will lead to waste of resources and cause tangible as well as psychological stresses to patients and their families. On the contrary, timely referrals lead to improved outcome of patient care. General Outpatient Clinic (GOPC) doctors refer patients to the AED for various reasons. They include: admission for further management, urgent specialist consultation (e.g. eye), urgent investigation (e.g. CT scan for suspected stroke), and treatment not available in GOPC (e.g., incision and drainage for large abscess). The management given at AED may not correspond to the aims of referrals, e.g. the patient is discharged from AED instead of being admitted into hospital ward. The difference between the outcome at AED and the aims of referral has been used as a measure of appropriate referrals.⁵

The aim of this study is to assess the appropriateness of the AED referrals of a public GOPC in terms of their outcome at AED.

Method

This was a retrospective study of the referrals to the AED by the doctors of one GOPC. The nursing staff kept hard copies of all referrals since October 2002. For the period between October 2002 and July 2003, all the referrals to AED were retrieved. From the Clinical Management System, the attendances of these patients at AED were traced and the referral outcome noted. Descriptive statistics, Chi-square test and intention-to-treat method (all those referred as denominator) were used to analyze the results.

Criteria taken as "appropriateness" for this study were (1) urgent hospital admission (2) urgent treatment not available at GOPC (3) urgent investigation not available at GOPC (4) urgent referrals to Specialist Outpatient Clinic (SOPD) (5) referrals that require further follow-up treatment by AED.

Results

There were 137 referrals to AED during the period. The age of the patients ranged from 23 days to 93 years with the mean of 57.5 years. There were 56.2% (77) female and 43.8% (60) male patients. Three patients defaulted attending the AED. The aims of the referrals included assessment, treatment, and investigation.

The specialties of referred problems included internal medicine, surgery, paediatrics, orthopaedics, eye and gynaecology, with internal medicine as the most frequently referred (51.8%). Table 1 summarizes the outcomes of the referrals.

The majority of referrals (102/137, 74.5%) required further attention at AED; hospital admission (81/137, 59.1%), urgent referral to other specialties (10/137, 7.3%) or AED follow-up (6/137, 4.4%). Our criteria of "appropriate" referral included five patients who required urgent treatment or investigation that were not available at GOPC (5/137, 3.6%) despite their discharge from AED (three CT brains; one for Plaster of Paris for wrist fracture; one for urgent electrolyte when the test was not available at GOPC); thus 74.5 % (102/137) were found to meet the criteria. If the four patients who defaulted or were discharged against medical advice were excluded, the rate of appropriate referral was 76.7% (102/133).

We performed further analysis of those "inappropriate" referrals (31/137 excluding defaulters and the DAMA patient). They can be classified into two categories:

1. Doctor factors


1. Incorrect diagnosis

Two patients were diagnosed or suspected to have papilloedema with hypertension (HT). The diagnosis was not confirmed and the patients were discharged after a period of observation.
One patient was suspected to have foreign body in eye. No foreign body was found after normal saline flushing.
One patient with suspected shoulder dislocation was discharged after a normal X-Ray at AED.
A two-month old baby presented with reduced feeding and irritability for one day. She was found to have oral thrush.

2. Inadequate knowledge/management skill

Two patients were found to have elevated haemstix without evidence of diabetic ketoacidosis. They were discharged after treatment at AED.
One 45-year old lady presented with menorrhagia and was found to have mild pallor. Blood test at AED found normal haemoglobin and clotting profile and was discharged with treatment.
One elderly with known hypertension defaulted follow up for a few months and was found to have high blood pressure (BP) (206/103 mm Hg, rechecked with no papilloedema) was referred to AED because of high BP and social isolation. She was discharged after a period of observation for BP at AED.
One eight-year old girl was referred because of mild to moderate asthmatic attack. She was discharged after nebulizer treatment.

3. Unfamiliarity with clinic facilities

One patient with foreign body in throat was referred with the reason of lack of equipment for direct laryngoscopy. However this was available at our clinic.

4. Lack of time/skill

Two patients were referred for incision and drainage of paronychia and infected sebaceous cyst. From the referrals we cannot determine whether the trainees referred due to lack of skill or time. The procedures should be within the scope of family physicians.

5. Uncertain diagnosis

One 77-year old lady without history of ischaemic heart disease presented with a four-day history of precordial discomfort and normal ECG at GOPC. She was diagnosed to have epigastric pain and was discharged after normal ECG, CXR and blood test. Another 50-year old with known HT presented with one-day history of chest discomfort and ECG finding of ST depression and T inversion in leads II, AVF, V1-6. However she was diagnosed to have chest wall pain at AED and discharged subsequently. Another 45-year old lady presented with non-specific chest discomfort for 10 days and on ECG was found to have ventricular bigeminy. She was found to have only ventricular ectopics at AED and was referred to a cardiac specialist with non-urgent appointment.

One 78-year old man presented with right eye redness for one day after being hit by an umbrella. He was referred because of "severe swelling and conjunctival haemorrhage". He was diagnosed to have subconjunctival haemorrhage and was discharged without further intervention.

One 20-year old young man presented with vomiting, vertigo and fever. He was referred for "? cause of vertigo". He was diagnosed to have gastroenteritis and was discharged after treatment.
One 81-year old lady was referred for suspected fracture of left elbow with no history of trauma. She was discharged after normal X-Ray.
Another 46-year old lady was referred because of suspected fracture patella after a fall. There was no fracture on X-Ray at AED and was discharged with the diagnosis of knee contusion.

2. Clinic factors


1. Lack of equipment/facilities

Four patients were referred because of high BP. They were presumably referred for observation although the reasons for referral were not stated clearly. Three were referred for abdominal pain and were suspected to have appendicitis. One with history of psychiatric illness presented with generalized urticarial rash for few days and chest discomfort for years. Presumably the doctor was suspecting angioedema. He was given intramuscular antihistamine and discharged after observation at AED. These eight patients might have been well managed at GOPC if a proper observation room together with appropriate staff were available. One patient with corneal abrasion was discharged home from AED with no further management. Another one with corneal abrasion was given routine referral to Eye specialist. These two referrals were considered as inappropriate by the pre-set criteria.

Discussion

This study shows that 22.6% of the patients referred to AED were discharged without further management plan. With explicit criteria in terms of outcome of the referral, i.e. admission, urgent referral to other specialties, AED follow-up and urgent investigation or treatment not available at GOPC, the appropriateness of the referrals from the GOPC studied was 74.5%.

From the analysis of the inappropriate referrals, the following strategies might improve the appropriateness of referrals:

1. Doctor factors


1. Staffing

A group of relatively stable manpower will ensure accumulation of experience. The current policy of contract and temporary staff will definitely increase the rate of inappropriate referrals. Our study period was nine months and more than 20 doctors had been working in our clinic.

2. Training

Training on referrals is also important. Formal teaching of common scenarios for referrals to new staff may improve their quality of patient care.

2. Clinic factors


1. Establishment of referral guidelines may allow easier access of up-to-date knowledge on case management and hence improve referral quality.
2. Improvement in clinic setting. Establishment of observation room and properly equipped minor operating room would facilitate family physicians to perform procedures that are within their capacity.
3. Improvement in consultation time. The current five to seven minutes consultation time discourages doctors to perform any procedure.
4. Screening of referrals by senior staff is useful. However, this may not be practical as many of the seniors also have heavy work load.
5. Communication with AED clinicians - Feedback is an important tool for learning. Currently AED doctors do not usually reply to us after the referrals. Regular feedback definitely facilitates family physicians' learning on referrals to AED. Regular meeting with AED clinicians to discuss cases could also be useful. We can also establish a database to record all the common inappropriate cases and be viewed by the new staff.

In another part of our study, we invited five family medicine trainers to judge the appropriateness of the referrals taking into account the working environment of our GOPC.7 They were asked to assess these referrals and see if they agreed with the appropriateness of the referrals or not, using a four-point scale:

1 - agree,
2 - doubtfully but tend to agree,
3 - doubtfully but tend to disagree,and
4 - disagree.

"Appropriateness" was taken as that "the patient under consideration should be referred".

They were blinded to the hypothesis of the study and the ratings of others. The result shows that 80.4% (59.9% rated 1', 20.5% rated 2') agree with the referrals. This may be explained by the fact that they had considered the working environment of GOPC doctors. From the raw data, the difference in the appropriate referrals between the trainers' judgment and the referral outcome does not reach statistical significance at p=0.05 (chi-squared test: c²=2.10, p=0.15, odds ratio = 0.71 [95% confidence interval: 0.45, 1.12]).

The methods of evaluating the appropriateness of referrals, as described in various studies, are largely subjective, without validation or international agreement. Family physicians, hospital specialists, and patients could have different opinions of appropriateness.⁸

Patel and Dubinsky in Canada used outcome measures similar to this study (admission, specialist consultation, intervention, investigation not available in family practice) and found that 75.5% of 196 patients referred to AED by family physicians were appropriate.⁵ The result is similar to this study. Fertig et al in the UK observed that 9.6% of general practitioners' referrals to hospital were judged as inappropriate by specialists while 15.9% inappropriate according to referral guidelines.⁴ Assessment by panel of experienced family physicians in another study revealed that 23% AED referrals could be treated in general practice,⁹ similar to our study.⁷ The only published local study that considered the appropriateness of referrals by family physicians was based on the judgment by a panel of nine family physicians according to an agreed model.³

The method of evaluation (outcome measures) and the result of this study are close to the prospective study by Patel et al in Toronto.⁵ However, we do not make any comparison of referral behaviours in the two studies due to different settings of family practice. In our study, the referrals were made by 22 doctors of variable experience. Some of them just finished their internship and were very new in family practice setting. Some of them just started working in our GOPC for only short period of time and might not know the clinic setting well. These certainly increased the rate of "inappropriate" referral.

Another limitation of our study is that the criteria of appropriateness are arbitrary. There are more factors affecting the decision by a family physician to refer. These include available equipment, e.g. properly equipped minor operating room, working hour constraint and lack of supporting staff, e.g. to observe the patient.

Our criteria regarded the following patients as inappropriately referred though under the working condition of the GOPC the referring doctors had few or no alternative options: two patients who had incision and drainage done for paronychia and infected sebaceous cyst respectively; one with intravenous saline infusion for gastroenteritis; at least three with elevated blood pressure were discharged after a period of observation; two patients were referred for urgent x-ray because of clinically suspected fracture.

Without validated and universally agreed method of evaluation, it is difficult to ascertain the appropriateness of referrals or to compare the findings of different studies. Further research on the measurement of this health-care outcome is much in need.

Conclusion

We describe an "outcome" method of evaluating the appropriateness of referrals by family physicians to AED. We find that 74.5% of the referrals are appropriate as judged by the defined outcome. Analysis of those inappropriate referrals indicates deficiency both in doctors and clinic setting which may be improved, with suggested recommendation.

Acknowledgement

We would like to thank Dr MH Ng (COS, AED, TKOH) and Dr CY Yeung (Resident, FM&PHC, UCH) for their assistance in this study.

Key messages

1. In a public general outpatient clinic, 74.5% of the referrals from a General Outpatient Clinic to an Emergency Department in the defined period were found to be appropriate by outcome measures at Emergency Department.
2. Analysis of inappropriate referrals showed modifiable contributing factors concerning doctors and clinic settings.
3. Referrals might be further improved with administrative measures, training and change of clinic settings.