The immediate discharge document
Section 2: The evidence base

2.1 Introduction

The evidence for this manual was collected in accordance with SIGN methodology.6 Literature searches covered the Cochrane Library, EMBASE, MEDLINE and CINHAL databases. A general internet search was also undertaken for relevant studies. The search covers the period from 1993 to 2001. The search was extended to include all types of study, as the topic of discharge documentation does not lend itself to clinical trials.

The vast majority of the studies reported were either audits, case control or cohort studies. Only one randomised controlled trial was identified.7 The lack of high level evidence has resulted in no highly graded recommendations.

In view of the paucity of the literature on discharge documentation, a letter was sent to the Chief Executives of all NHS Trusts in Scotland inquiring about specific new or innovative practice in this area. A number of relevant and innovative approaches were identified and their originators were invited to present their work to the guideline development group.

2.2 Evidence from audit

Since the publication of the original document in 1996,1 several publications have indicated continuing problems with the production of discharge documentation.8, 9, 10 The problems relate to content, the process of production and transmission.

A prospective audit of discharge summaries using a computerised system of discharge letter generation reported a 10% overall error rate, 22% of which were errors relating to the diagnosis.10 Another prospective audit scored 100 ‘interim’ discharge letters on their level of completeness and found an average score of 13 out of a potential 16. Only 83% were legible and 30% did not name the doctor preparing the summary. Only 43% were dispatched within five days from discharge.9

One audit has revealed differences between the ward prescription and the discharge prescription in 13% of cases. Differences between the ward prescription and the final letter were identified in 29% of cases.11 Fifty seven percent of patients had medication-related problems two weeks after discharge and some summaries did not reach the GP.11

2.3 The content of the dataset

The original document argued the importance of a minimum dataset as a basis for system specification with a view to electronic data transfer. It also specified such a dataset, emphasising that it was only a minimum. A minimum dataset should be evidence based and include all those items seen as essential. A number of papers have examined what doctors, particularly GPs, perceive as being the important data items.12, 13, 14, 15, 16, 17

A study investigating the influence of various factors on the quality of discharge summaries identified content as the most important factor in determining the quality of a summary.12 Content was classified into preadmission information, hospital information, and discharge information. Discharge information was the most important determinant of quality, followed by hospital information. Evidence level 3

The top ranking items were:

Other studies13, 14, 15 show similar results. These items are a minimum requirement and there is a potential for including much more data in a discharge document. Evidence level 3

In a study designed to assess the value of including a detailed functional assessment measure in a discharge document, 61% of general practitioners found such detailed information moderately useful and 28% found it very useful.16 In contrast, 44% of consultants receiving the same type of information found it moderately useful and 48% found it very useful. This may indicate that too much information can be included in a discharge document. A second study reported that conciseness is valued and suggested using a smaller size of paper.13 Evidence level 3

The variation in the perceptions of important data items between professional groups has also been investigated.12, 15, 17 Broad agreement between hospital and family practitioners on the relative importance of major data items has been reported. Significant differences were identified in the ratings given to continued care, prognosis, psychosocial factors and functional ability at discharge.12, 15, 17 These differences should be appreciated by hospital staff preparing discharge documentation to ensure the resulting document fulfils the needs of the recipient. Evidence level 3

2.4 Who should prepare the IDD?

In current hospital practice the completion of discharge documentation is often delegated to the more junior medical staff.18 Audit has shown that senior staff produce discharge summaries with fewer errors.10 It has been estimated that a Senior House Officer (SHO) producing 20 final discharge letters per week may need up to four hours to complete this task with the consequence that other duties take priority.18 As a result letters are often sent out late. No evidence was identified that explored who was the best person to complete the IDD. The notes accompanying item 23 of the dataset give more details of who may complete the document.

2.5 Format and completion of the IDD

The majority of family doctors prefer a structured discharge document rather than one with a narrative format.7, 13, 19, 20, 21 The reasons for this preference include completeness, readability, conciseness and the ease of locating key information. Despite these perceived advantages, a study looking at the influence of headings and structure on the readability and retention of data was unable to show that headings improved either factor.22 Evidence level 2+

Structured format discharge documents can easily be produced by computer systems. A randomised controlled trial comparing dictated discharge summaries to computer generated summaries found that computer generated summaries were shorter than dictated summaries and were produced more quickly.7 The computer-generated summaries contained data relating to active past medical history and medications more often than the dictated summaries but contained admission diagnosis and discharge functional status less often. A global assessment of quality found no difference between the two types of summary.7 Evidence level 1+



2.6 Who should receive the discharge document?

Discharge documentation serves many functions including the prescription of medicines, general communication and serving as a record of admission. Specialties such as obstetrics require more detailed information than is contained in the core dataset. The postdischarge requirements of all patients vary considerably, ranging from a discharge prescription for short term analgesia with no postdischarge review, to a comprehensive discharge plan with contributions from and for, for example, medical, nursing, occupational therapy and social work staff.

In many hospitals the discharge document acts as a prescription. Discrepancies between hospital and community prescriptions have been recorded.11 A cohort study examining this problem found that sending a copy of the discharge document to a community pharmacist reduced the discrepancy rate from 52.7% to 32.2%.23 Evidence level 2+

No evidence was identified addressing the wider distribution of the IDD. It is recognised that many people in addition to the patient’s family doctor might wish to receive a copy of the discharge documentation including community pharmacists, other GPs and community nurses. Potentially complex arrangements require the involvement of carers and the necessary resources for optimal implementation. The IDD is not only for general practitioners.

The IDD is a form of communication for use between professionals but patients do have the right to see it. Patients should be given the option of having a copy of the IDD, although the production of separate patient-centred information may be more relevant. The production of patient-centred material is outside the remit of this document.3


2.7 How should the discharge document be transmitted?

The time from discharge to receipt of document is a major cause of concern, with some discharge documentation not reaching the intended recipient.11 Delays in the receipt of the document can lead to potential problems, including errors in prescribed drugs.11 Evidence level 3

A cohort study investigating discrepancies between hospital and community prescriptions found that discrepancies can result from slow or non-delivery of discharge letters, but can also arise from errors made within the hospital.11, 23 Three studies have investigated the advantage of fax transmissions of the discharge document and found that its use resulted in quicker transmission of a usually more legible, discharge documentation.24, 25, 26 Faxed transmission of discharge documents does speed delivery but carries the risk of transmission to an incorrect number. This can be minimised by using machines with preprogrammed abbreviated dialling codes. Current Scottish Executive Health Department guidance says that the use of fax transmissions for the transfer of personal health information in NHSScotland should only be performed with extreme caution.27 Evidence level 2+,4

With the increasing use of computers to produce discharge documents there is the possibility of direct electronic transfer that should offer the same advantages of fax transmissions while reducing the need for the transcription of prescriptions.28

[Contents][Back][Top] [Next]

Scottish Intercollegiate Guidelines Network
Home