Scottish Medical Journal

D A Nunez , Consultant/Clinical Senior Lecturer , Department of Otolaryngology , Southmead Hospital , Bristol BS10 5NB

Abstract

Hospital and general practitioners should establish good communications so that continuity of care can be maintained when patients are discharged from hospital. A discharge letter was routinely prepared utilizing the standardized morbidity registration forms by the house surgeon for delivery to the patient’s general practitioner. The authors developed a new discharge letter, which was sent under separate cover to the general practitioner. The two discharge letters were assessed utilising data fields recommended by the Scottish Intercollegiate Guidelines Network (SIGN).

162 consecutive paediatric patient discharges were identified. The mean number of items present in the existing and the new discharge letters were 13.7 + 1.8 and 12.2 + 3 respectively. The new discharge letter met the SIGN guidelines more closely than the existing discharge letter.

Introduction

Good communication is vital in delivering the best quality of health care. In order to facilitate communication, discharge letters are either sent to the general practitioners by post or given to the patients for hand delivery to their doctor.

It is usual practice in the United Kingdom to prepare a discharge letter. This most often takes the form of a multiple non-carbon copy form with pre-printed headings, completed by hand and either given to the patient for delivery to their general practitioner or mailed. This letter also known as the immediate discharge letter takes about 2 days to arrive at the general practitioners’ surgeries.1,2 It seems to be an efficient way of communication, however the handwriting can be illegible and some general practitioners find these letters lacking in personal touch.3

In some clinical units, a further typewritten letter is mailed to the general practitioner once the patient is discharged from hospital. This letter is usually dictated by a more senior doctor and can contain more detailed information concerning diagnosis and treatment than the immediate discharge letters.

General practitioners prefer discharge letters to be written using a problem-orientated approach,4 well structured, concise, typewritten and delivered to them with minimal delay.5-8 Sometimes the patients visit their general practitioners soon after they have been discharged from the hospital. They may want to discuss the future management plan of their illnesses, results of their investigations and require drug prescriptions. The general practitioners will be less able to satisfy these demands if they have not been informed of the progress of their patients.

The content of the discharge letter including the administrative and clinical details, which general practitioners prefer have been studied.8-10 Ideally, the immediate discharge document should provide authoritative diagnostic and treatment data and be coded to allow clinical audit.11

A minimum data set (table 1), consisting of 21 items, for the immediate discharge document was developed by a steering committee under the aegis of the Royal College of Surgeons of Edinburgh, the Royal College of Physicians of Edinburgh, the Royal College of Physicians and Surgeons of Glasgow and the Royal College of General Practitioners (Scottish Council). This minimum data set was recommended for use in Scotland by the Scottish Intercollegiate Guidelines Network (SIGN). The SIGN recommended that the immediate discharge document should be read and approved by senior doctors, completed before the patient is discharged and it could be the sole discharge document. Development and implementation of local protocols were encouraged. It was suggested that audit committees should set up studies to evaluate the change of practice and to establish its effectiveness.11

A Scottish Medical Record form completed by a junior doctor on the day of patient’s discharge was the only letter used when a child was discharged from the otolaryngology service prior to 1995. One author (DAN) at the end of 1995 initiated the practice of a second type-written discharge letter dictated by a senior doctor which has become the more common practice in the local paediatric otolaryngology service. The aim of the audit was to assess the extent to which the pre-existing and the second discharge letters met the selected criteria laid down by the Scottish Intercollegiate Guidelines Network (SIGN).

The Scottish Medical Record forms were completed by hand by a junior doctor on the day of the patients’ discharge from the hospital. One copy was sent to the general practitioner by first class post as soon as the patient was discharged. Further copies were sent to the hospital pharmacy for dispensing medication and to the medical records office. A final copy was filed in the patient’s hospital case note.

A specialist registrar or consultant dictated the second discharge letter. The letters were written in a problem-orientated approach. They were usually dictated on the day before or the day when the patients were discharged. Occasionally the letters were dictated after the patient had left the hospital. For patients who were admitted for routine operations, the discharge letters were dictated immediately after the surgery in the operating theatre. The letters were typed by the secretary, signed by the doctors and then sent to the general practitioners by first class post.

The clinical records of otolaryngology patients who were admitted to the Royal Aberdeen Children’s Hospital under the care of two consultant otolaryngologists between 1 January and 30 June 1997 were studied. General practitioners (GPs) in the Grampian region were invited to participate in the study. Patients whose general practitioners did not want to participate were excluded. Patients from the Orkney and Shetland Island Health boards were excluded due to the impracticality of visiting the GP surgery to collect the required information. Finally patients who had changed their general practitioners and who could not be traced were also excluded. A data collection sheet was designed and information was collected from the hospital’s and GP’s notes. The date the discharge letters arrived at the GPs surgeries was recorded.

The Chi-squared test or Fisher’s exact test were used to statistically compare the proportions of the existing and newly implemented discharge letters that contained 16 of the 21 data field items recommended in the SIGN guidelines (table 1). The ‘patient’s name’, ‘date of birth’ and ‘hospital number’ were collectively analysed under one heading ‘patient’s identification’ as these items were invariably present together. The ‘operations/procedures’ data field was not included in our assessment because not all patients who were admitted required operations. ‘Results Awaited ‘ and ‘Further letter to follow?’ were not analysed because by the nature of the study design the new discharge letter was more likely to have the additional results and also unlikely to state ‘further letter to follow’.

Recommended by SIGN

Hospital
Patient’s Identification
General Practitioner’s Name
Consultant’s Name
Ward/Department
Date of Admission
Date of Discharge
Reason for Admission / Transfer
Mode of Admission
Main Condition
Other Active Problems
Operations / Procedures
Medication on Discharge
Allergies
Other Plans on Discharge
Comment
Results Awaited
Further letter to follow?
Read & Approved
Contact Name
Signed

Data Set Analysed

Hospital

Patient’s Identification

General Practitioner’s Name

Consultant’s Name

Ward/Department

Date of Admission

Date of Discharge

Reason for Admission

Mode of Admission

Main Condition

Other Active Problems

Medication on Discharge

Allergies

Plans on Discharge

Comment

Signature

The new discharge letter was prepared and signed by a registrar or consultant, which is at variance with the proposal of the SIGN committee, which anticipated that the more senior clinician would only read and approve the discharge letter. The new discharge document protocol developed required the senior clinician to actually prepare the document and thus the data field ‘read and approved’ became redundant for analysis purposes in this study.

Students t test was used to analyse the difference in the mean times taken for the two discharge documents to reach the GP.

One hundred and sixty two patients were admitted to the Royal Aberdeen Children’s Hospital between 1 January and 30 June 1997 under the care of two consultant otolaryngologists. One hundred and thirty general practitioners in the Grampian area were identified of whom 123 agreed to participate in the audit, a participation rate of 94.6%. A total of 160 (98.8%) existing and 152 (93.8%) newly implemented discharge letters were studied.

The mean number of items present in the existing discharge letters, were 13.7 + 1.8. ‘Hospital’s name’, ‘patient’s name’, ‘date birth’ and ‘hospital identification number’ were always present (100%). The frequency of ‘general practitioner’s name’, ‘consultant’s name’, ‘date of admission’, ‘date of discharge’, ‘reason for admission’, ‘ward’, ‘plans on discharge’, ‘comment’ and ‘signature’ present in the letters varied from 92.6% to 98.8%. ‘Main condition’ appeared in 78.1% of the total letters. ‘Mode of admission’, ‘other active problems’ and ‘medication on discharge’ were present in 1.2 to 18.75% of the letters. ‘Allergies’ was not present in any of the letters. The mean time taken for the existing discharge letters to arrive at GPs’ surgeries was 4 + 5.7 days (range 0 to 34 days).

The mean number of items present in the newly implemented discharge letters was 12.2 + 3.2. ‘Hospital’s name’, ‘patient’s name’, ‘date of birth’ and ‘hospital identification number’, ‘general practitioner’s name’ and ‘signature’ ‘Ward/department‘ were always present (100%). The frequency of ‘consultant’s name’, ‘date of admission’, ‘date of discharge’, ‘reason for admission’, ‘main condition’, ‘plans on discharge’ and ‘comment’ present in the letters varied from 82.2% to 99.3%. ‘Mode of admission’, ‘other active problems’, ‘medication on discharge’ and ‘allergies’ were present in 0.7% to 22.4% of the letters. The mean time taken for the new discharge letters to arrive the GPs’ surgeries was 16 + 7.9 days (range 10 to 50 days).

The statistical package for the social sciences version 9.0 was used for analysis. The proportion of existing and newly implemented letters containing data in each of the 16 fields chosen for analysis were compared by c² test or Fisher’s exact test (when the expected frequency in any cell was less than 5). Multiple comparisons increase the likelihood of a spurious statistically significant result being found and hence statistical significance has been set at 0.003 for this study rather than the usual 0.05 (Bonferroni correction).

The mean number of items present in the existing and the new discharge letters were 13.7(standard deviation, s.d.; + 1.8) and 12.2 (s.d., + 3) respectively. The ‘hospital’s name’, ‘patient’s name’, ‘date of birth’, ‘hospital identification number’, ‘general practitioner’s name’, ‘ward/department’ and ‘signature’ were almost always present (94.4 to 100%) in both discharge letters equally. Other items including ‘consultant’s name’, ‘date of admission’, ‘date of discharge’, ‘reason for admission’, ‘main condition’, ‘plans on discharge’ and ‘comment’ were present variably (78.4 to 99.3%). There was no statistically significant difference in the frequency with which the reason for hospital admission was stated in the existing compared with the new discharge letter, 93.8% and 99.3% respectively (c² test, p = 0.007). The patient’s main condition was however less commonly mentioned in the existing than in the new discharge letter, 78% and 93% respectively (c² test, p < 0.001).

The remaining 7 items including ‘ward’, ‘mode of admission’, ‘other active problems’, ‘medication’, ‘allergies’ were rarely (0 to 22.4%) present. There was no statistically significant difference between the newly implemented and the existing discharge letters with respect to these data fields.

There was a difference in the mean times for the GPs to receive the two discharge letters. The mean time taken for the existing and the newly implemented discharge letters to reach the general practitioners were 4 (s.d., + 5.7) days and 16 (s.d., + 7.9) days (Student’s t test, p < 0.001).

The interface between the hospital and the community is important in clinical practice. Many studies had been carried out to assess how to improve communication between clinicians in hospitals and those in the community.1,2,3,5,6,8,9 A new discharge letter was introduced to our clinical practice. This study evaluates the change of practice utilising criteria recommended by the SIGN on the interface between the hospital and the community.

Both the existing and newly implemented discharge letters have their advantages and disadvantages. The existing discharge letters were handwritten and were quick to prepare. They were completed and sent to the general practitioners on the day of the patients’ discharge. The existing letters took an average of 4 days while the newly implemented letters took 16 days to arrive at the general practitioners surgeries. The existing discharge letters, therefore, required a significantly shorter time to prepare and caused minimal delay.

The newly implemented discharge letters were prepared by a more senior clinician, either an intermediate grade or consultant surgeon. They were dictated the day before, or soon after the patients were discharged from the hospital. In patients admitted for elective surgery, the new discharge letters were dictated after surgery in the operating theatre. By design the newly implemented letter would arrive later than our existing discharge letter. Bowie et al, identified four periods of time causing delay in the delivery of discharge letters.12 Firstly, the time taken from the patient’s discharge to letter dictation. Secondly, the time taken from dictation to typing. Thirdly, the time from typing to signing the letter. Finally the time from signing to the letter being mailed.

The newly implemented discharge letters more often stated the main condition diagnosed than the existing discharge letter. In addition the new discharge letters were type-written in a problem-orientated approach which the general practitioners preferred ¹³.

The disadvantage of the delay in the arrival of the new discharge letter at the GPs surgery arose for all the reasons previously outlined in the literature. Emergency admissions were the most common source of delayed discharge summaries as the letters were not dictated until results of all investigations were available after the child was discharged. There was a delay in the case records getting to the secretary’s office after discharge. The Children’s hospital was not the base hospital for any of the doctors involved in preparing the new discharge summary, introducing another element of delay. An electronic patient record would surmount the majority of these sources of delay. A secure electronic network would allow access less restricted by the location of the clinician to the results of investigations prior even to the dispatch of written laboratory reports. A significant short-circuiting of the inherent delay built into our current dependency on paper reports and the case notes would be achieved. It may also offer an opportunity for electronic editing of the dictated discharge summary by the clinician and a faster route than the postal route to send the resulting document to the general practitioner. Electronic patient record systems are already in use in some NHS units and the advantages described are feasible.

The authors thank the Grampian Central Audit Committee for providing financial support, General practitioners in Grampian for their cooperation and Janet Rollo for data collection.

Category	Existing letters (n = 160)	New Letters (n = 152)	p values (<0.003)
Hospital’s name	160 (100%)	152 (100%)
Patient’s identification	160 (100%)	152 (100%)
General practitioner’s name	156 (97.5%)	152 (100%)
Consultant’s name	156 (97.5%)	148 (97.4%)
Date of admission	158 (98.8%)	143 (94.1%)
Date of discharge	148 (92.6%)	144 (94.7%)
Reason for admission	150 (93.8%)	151 (99.3%)
Plans on discharge	154 (96.3%)	150 (98.7%)
Comment	148 (92.5%)	125 (82.2%)
Signature	152 (95%)	152 (100%)
Ward/Department	151(94.4%)	152 (100%)
Main condition	125 (78.1%)	142 (93.4%)	<0.001
Other active problems	29 (18%)	34 (22.4%)
Medication	30 (18.75%)	16 (10.5%)
Mode of admission	2 (1.2%)	2 (1.3%)
Allergies	0 (0%)	1 (0.7%)

Scottish Medical Journal

Audit of a change in otolaryngology discharge letters using the Scottish Intercollegiate Guidelines Network (SIGN) recommendations.

Abstract

Introduction

Back to October Index