General practitioner understanding of abbreviations used in hospital discharge letters

Deficits in discharge letters can contribute to a failure of information transfer. Studies have found high rates of omissions and errors in such letters.2-4 This contributes to errors in care after discharge. One study found that 49.5% of patients discharged from a large academic medical centre experienced at least one medical error relating to change of care on discharge.2

In this article, we focus on the potential danger of using abbreviations (shortened forms of words or phrases5) in medical communication. Abbreviations used in medical communications are either acronyms or initialisms. Acronyms use the initial letters of words and are pronounced as words (eg, ASCII, NASA); initialisms use initial letters pronounced separately (eg, BBC).5 Abbreviations are commonly used in medical specialties, but may not be understood by the broader profession. Doctors are under pressure to complete discharge letters in a timely fashion, and abbreviations may be used to facilitate this process.

We identified few published studies of the frequency of abbreviations in discharge letters.6,7,8 Some reported that abbreviation use is increasing and identified this as a concern. A recent audit at Royal Melbourne Hospital reported that 20.1% of all words in discharge letters were abbreviations.8 Another study audited abbreviation use in inpatient medical records and surveyed members of an inpatient multidisciplinary team for their understanding of abbreviations.9 The mean correct response rate was 43%, with Postgraduate Year 1 doctors posting the best scores (57%) and dietitians posting the worst (20%).

However, we identified no published studies determining whether the abbreviations used in hospital discharge letters are understood by GPs, who are usually the recipients of discharge letters.

Methods

We retrospectively analysed 200 electronic hospital discharge letters (eDLs) of patients discharged from Nepean Hospital, Sydney, a tertiary referral centre, from 31 December 2012, working backwards to 18 December 2012. We stopped at this point because few new abbreviations were being identified. To be included in the audit, an eDL had to be addressed to a GP.

We chose 31 December to begin the analysis to provide a representative sample of junior doctors who had a minimum of almost a year of hospital experience.

The meaning of each abbreviation was inferred from the surrounding text, and abbreviations were categorised as shown in Box 1.

Survey of GPs

From the audit, we developed a survey using the 15 most commonly used abbreviations plus five less frequently used but clinically important abbreviations. We determined that abbreviations of investigations, management or services were likely to be most clinically significant, based on our clinical experience and the potential consequences of misinterpretation. We defined commonly used abbreviations as those that were used at least 20 times in the audit. In the resulting survey of GPs, each abbreviation was provided in the context of a phrase in which it had been used in a discharge letter (Appendix).

To provide adequate precision, we aimed for 100 GP responses. The survey was mailed to all 240 GPs listed in the 2014 edition of the Medical Practitioners’ Directory for the Nepean, Blue Mountains and Hawkesbury areas. This was the most extensive directory of GPs in this area available to us. Responses were returned in a coded envelope inside a postage-paid envelope. GPs who did not respond were resent surveys on up to two additional occasions.

Outcome measures

Survey responses were analysed to determine what proportion of GPs understood each abbreviation.

Ethics approval

The study was approved by the Nepean Blue Mountains Local Health District Human Research Ethics Committee.

Results

Electronic discharge letter audit

We found 321 different abbreviations in the 200 eDLs audited (a rate of 1.6 new abbreviations per eDL and 7.1 total abbreviations per eDL); most were initialisms. The frequency of abbreviations in eDLs is shown in Box 2.

Hospital coding-approved abbreviations accounted for 62.6% of all abbreviations identified. Seven unapproved abbreviations (2.2%) were in common use (ie, found more than 20 times in the audit).

GP survey

The response rate was 55% (132 of 240 GPs). No abbreviation was correctly interpreted by all GPs, but 10 abbreviations (50%) were interpreted correctly by 97.0% of GPs (128).

The frequency of incorrect interpretation of all abbreviations in the survey is shown in Box 3. Box 4 shows the range and frequency of individual GP scores.

Discussion

The results of our survey show that there is poor understanding among GPs of abbreviations used in hospital discharge letters. The response rate to our survey was fair, so our results are likely to be representative of GPs in the area.

Worryingly, more than half of the abbreviations we found related to investigations, management or services that we considered to be the most clinically significant categories. Misinterpretation of abbreviations by GPs can adversely affect patient care through duplication of investigations, failing to institute treatment based on investigation results or failing to follow up with recommended management. We could find no studies that identified which types of abbreviations confer the worst outcomes if misinterpreted. Also of concern is that almost half of the abbreviations we identified were used only once in the 200 eDLs.

The difference identified in the use of abbreviations by junior doctors and understanding of abbreviations by GPs suggests a lack of consistency between the language commonly used in hospitals and that used by GPs. It is uncertain how well understood these same abbreviations are by hospital doctors in different specialty areas. The language of abbreviations may also vary between hospitals. Common abbreviations found previously in Royal Melbourne Hospital discharge letters8 were different from those we found. The five most common inappropriate ambiguous or unknown abbreviations in the Royal Melbourne Hospital audit were not found in any eDL in our audit. Their abbreviation rate was higher, with a mean of 10.5 new abbreviations per discharge letter compared with our rate of 1.6. Widespread use of abbreviations in paediatric medical notes with no standardisation and difficulty in interpretation by health care professionals has also been previously reported.11

Our study has some limitations. Non-responding GPs might have scored differently on the survey compared with those who responded. Also, we did not ascertain GP demographic characteristics such as length of career outside the hospital setting. GPs with more recent hospital practice may better understand these abbreviations. In addition, we could not assess GPs’ understanding of most abbreviations we identified in the eDL audit because of the large number identified. However, we expect that understanding of these less frequently used abbreviations would be poorer than for the 20 we included in our survey. Also, this study was conducted in a single centre, so the results may not be generalisable to other centres. However, junior doctors are drawn from many universities and it is likely that discharge practices are similar in other hospitals.

Conclusion

Discharge letters are an essential means of communication between hospitals and GPs to facilitate optimal care of patients when they return to the community. All abbreviations used should be understood by all GPs. Strategies to improve communication by means of discharge letters are urgently needed. Potential solutions include banning the use of abbreviations in eDLs or using only a limited number of hospital-approved abbreviations and providing GPs with an approved abbreviation list. Another option would be use of computer software to auto-complete mutually exclusive abbreviations (ie, allowing only one possible meaning for each).

4  Proportion of general practitioners receiving particular survey scores for correct interpretation of abbreviations

---