Study objective Patients discharged from the emergency department (ED) should be informed comprehensively and accurately about the diagnosis, future examinations and follow-up care. This study investigates: (1) how comprehensively patients are informed by physicians on discharge; (2) how accurately patients remember this information after discharge; (3) how well informed overall patients leave the ED; and (4) whether informedness relates to patient satisfaction.
Methods This study compares: (1) information given during discharge conversations, based on audio recordings of the conversations, with (2) accuracy of patient recall of this information, based on postdischarge interviews. During these interviews, the authors also assessed (3) amount and accuracy of information provided during treatment. Furthermore, the authors obtained (4) satisfaction ratings by physicians and patients. Data were collected for 96 patients during 20 shifts.
Results Sufficient information was provided in 83% of discharge conversations. Patients correctly recalled 82% of information received about diagnosis, 56% about examinations planned and 72% about follow-up treatments. Information related to medication was most prone to forgetting or distortion. Altogether, 43% of the patients left the ED correctly informed about diagnosis, planned examinations and follow-up. Patient satisfaction ratings were high (mean 4.7 on a 5-point Likert Scale) and not related to informedness of the patient.
Conclusions Patients had important information deficits when leaving the ED, and information transmission needs to be improved. The physician–patient discharge conversation seems an ideal opportunity for enhancing patient informedness. Standardisation of discharge procedures and training physicians in how to ensure that patients actually understand the information provided are needed.
- Discharge information
- information retention by patients
- physician–patient communication
- data management
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
- Discharge information
- information retention by patients
- physician–patient communication
- data management
Providing information to patients is a significant part of care in general;1 this also applies to emergency departments (ED). After a visit to the ED, patients should leave with an understanding of their illness and with accurate information about future examinations and follow-up care. However, previous studies have shown that many patients have important information deficits after ED discharge,2 ,3 which may lead to low adherence to instructions.4 In addition, patient satisfaction is related to communication aspects and information received,5–7 with patients generally preferring more information than they receive.8 Good patient information can improve treatment options,9 reduce patient insecurity and strengthen trust in the patient's own ability to take action after discharge.10 The better the patients understand what their disorders entail and what measures are required, the more likely they will be able to comply with the prescribed administration of medication and follow-up appointments.11
In the ED under study, written discharge information is reserved for a few specific cases (eg, ankle injuries). Typically, information is transmitted from physicians and nurses to the patients throughout their stay, and a final physician–patient discharge conversation takes place before patients are released. This discharge conversation aims at providing patients with a summary of the diagnosis, examinations performed and planned, and follow-up therapy. If medication is prescribed, information about the purpose, dosage, when to take it, duration of treatment and possible side effects should be provided. However, the situation in an ED is characterised by time pressure, frequent interruptions, noise and the lack of an established physician–patient relationship; patients are often particularly stressed. These factors threaten effective communication, and previous studies have shown that the information provided at the end of an emergency consultation is often insufficient and incomplete.2 ,12 ,13
Although information provided by caretakers is pivotal for patient informedness, other aspects influence the amount and accuracy of information retained by patients. First, information is prone to be forgotten, and patients often do not correctly understand the information provided to them.3 ,14 It is not the information communicated but the information understood that is the decisive factor for compliance.15 Second, the physician–patient discharge conversation is not the only source of information; patients also receive information during treatment. Thus, in order to correctly assess patient informedness, we have to take into account: (a) the information provided in the discharge conversation, (b) how much of this information is correctly recalled by the patient and (c) how much information the patient recalls from other sources than the discharge conversation.
Goals of this investigation
Most previous studies have assessed patient informedness based on surveys rather than on actual behaviour, making it impossible to determine to what extent lack of information is due to the information not having been provided versus information provided but not correctly remembered. The current study therefore uses three sources: (1) audio tapes of the physician–patient discharge conversation; (2) standardised interviews with the patients after the discharge conversation; and (3) information contained in the patient file. We thus could assess the comprehensiveness of the information provided and determine how much of this information patients remembered correctly. Information reported by patients but not discussed in the discharge conversation was checked for accuracy against patient files. Altogether, this method allows estimating patient informedness at time of discharge and the extent to which lack of information was due to this information not being provided versus not being (accurately) remembered.
As a secondary outcome, we assessed patient and physician satisfaction with the information transmitted and the association of these satisfaction measures with objective information transmission.
Setting and participants
The study took place at a university hospital ED with approximately 30 000 consultations per year. During 20 randomly selected days in August and September, 2007, 14 late shifts (15:00–23:00), four early shifts (07:00–15:00) and two night shifts (23:00 to 7:00) were observed. We focused on late shifts because patient admission is the highest during this time.
Inclusion criteria for patients was their basic ability to communicate (conscious; able to answer; speaking German, French, Italian, English or Spanish). Eligibility was assessed by the triage nurse on patient admission, and eligible patients were approached by an external researcher (a psychologist not related to the ED) who introduced herself to the patient (normally in the waiting area), and asked the patient to participate in a study about patient satisfaction. She provided detailed information about the study protocol at this point. In order to avoid patient reactivity and a potential bias, the specific goal of studying information transmission was not mentioned. Patients who consented to participate and who were discharged from the ED during the shift (either home or transferred to a ward within the same hospital) were interviewed after discharge. The researcher highlighted the patient's name in the electronic record to alert attending physicians to audio taping the discharge conversation.
Study design, methods of measurement and data collection
The study protocol was accepted by the ethics board of the hospital. (1) The discharge conversation with the patient was audio taped, allowing for the assessment of content and amount of information given by the physician. (2) About 5–10 min after discharge, the patient was interviewed by the external researcher who had asked for consent to the study. The time frame was chosen to minimise memory loss effects over time. During the standardised interview, patients answered questions about the diagnosis, further examinations planned and follow-up treatments. They were also asked whether they had received a prescription for medication, and if yes, about medication details (name of drug, purpose, when to take it, how long to take it and possible side effects). (3) Both the physician and the patient rated their satisfaction with the information transmission during the discharge conversation on a 5-point scale (1=‘very dissatisfied’; 5=‘very satisfied’).
Standard for discharge conversations
Two physicians and two nurses from the ED served as experts for defining the standard for information that should be discussed with the patient on discharge. These included: (i) information about the main diagnosis (disorder, injured organ); (ii) main examinations performed in the ED (eg, blood tests, X-rays, CT scans); (iii) further examinations planned (eg, additional CT scans); and (iv) follow-up therapy (eg, monitoring by specialists, changing dressings). If medication was prescribed, information should be provided regarding its name, purpose, when to take it, how long to take it and possible side effects.
Coding of physician–patient discharge conversation
The discharge conversation was transcribed word by word and then coded with regard to the standards described above. Coding categories included diagnosis, examinations performed, examinations planned and follow-up care, and medications prescribed. For medications prescribed, we coded whether the name, purpose, use, treatment duration and side effects were mentioned. Coding categories were 1 (discussed), 0 (not discussed) and 9 (does not apply, eg, if no medication was given); the content of the information was summarised.
Assessing information retained by the patient
Postdischarge interviews were transcribed word by word. Coders identified information mentioned by the patient for the categories mentioned above. We then compared the information recalled in the interview to the content of the discharge conversation, thus assessing completeness and accuracy of information retained. For each category, it was coded whether (1) or not (0) the patient correctly remembered the information provided by the physician.
To assess information from sources other than the discharge conversation, information mentioned by the patient during the interview but not discussed in the discharge conversation was checked against the patient file to assess its accuracy.
Fifteen per cent of the discharge conversations and the interviews were coded independently by two researchers. Inter-rater reliability was assessed by Cohen's κ, which was high (κ>0.76 for all categories). Three researchers assessed accuracy of information only mentioned during the patient interview against the patient file; there were no disagreements.
The medical and nursing experts determined four of the five aforementioned topics as standard for comprehensiveness of the information provided to the patient: (a) diagnosis, (b) examinations performed, (c) examinations planned and (d) follow-up treatment. To establish comprehensiveness score of the patient–physician discharge conversation, one point was assigned for each of these topics being discussed. Four points were considered indicating ‘good’, three points ‘sufficient’, two points ‘insufficient’ and one point ‘completely insufficient’ information transmission.
Patient informedness index
To assess the level of information accurately retained by the patient, a similar index was established for the postdischarge interview. One point was given for correct mentioning of: (a) diagnosis, (b) planned examinations and (c) follow-up treatment. Examinations performed were not included in this index, because it is not of immediate relevancy for the patient. The informedness index ranges from 0 to 3: 0=patient knows nothing, 1=knows one, 2=knows two and 3=knows all three domains. Note that this index refers to all sources of information, not only information given during the discharge discussion.
Data analysis was done with SPSS V.19, a commercially available statistical software package.
During 20 shifts, 318 patients were admitted to the ED. Of these, 209 (66%) were eligible to participate: 109 (34%) were excluded based on the inclusion criteria and 67 (21%) declined to participate. Of the 142 eligible patients who agreed to participate, 42 could not be included due to organisational, medical or technical reasons (eg, sudden change in condition, delay of discharge); in four cases, physicians failed to record the discharge conversation. The final sample therefore is 96 patients. Their mean age was 47.5 years (SD: 19.09), 29 (30%) were female subjects. For 93 of these patients, questionnaire data from the physician were provided, and patient interviews were conducted for 89 patients (two patients withdrew their initial consent after the discharge conversation; three patients left the emergency room without known reason before the interview could be conducted; one patient felt too much discomfort to complete the interview, and one patient received such a severe diagnosis that the interview was cancelled for ethical reasons). Results reported are based on the maximum of data available for each analysis.
Physician–patient discharge conversation: information discussed and information correctly retained by the patient
The mean duration of the physician–patient discharge conversation was 4:01 min (SD 2.40 min); the median was 3:23 min (range 20 s to 15:48 min).
The first column of table 1 shows topics discussed during the discharge conversation. Most physicians informed about the diagnosis (90%) and about follow-up treatments (99%); examinations performed and follow-up treatments were mentioned in about half of the discharge conversations. To patients who received medication (n=71), physicians mentioned most often the name (76%), purpose (64%) and potential side effects (62%) of the drug(s) involved. Specifics about the use of the medication were discussed with 56% of the patients; but only 30% of the patients were given information about how long they should take medication.
The second column of table 1 shows what percentage of information provided during the discharge conversation was correctly recalled during the patient interview. The most accurately remembered information was related to the diagnosis (82%) and to follow-up treatment (72%) as well as to the purpose of drugs prescribed (64%). However, almost half of the patients failed to remember correctly what examinations were planned; only 30% correctly remembered medication side effects that the physician had informed them about. Note that all percentages refer to the information given, not to the number of conversations or interviews.
Informedness of patients after discharge
The physician–patient discharge conversation is not the only source of information the patients drew upon. The third column of table 1 takes this fact into account and shows the percentage of patients who reported correct information about diagnosis, treatment and medication, regardless of source. Ninety per cent of the patients held correct information about the diagnosis; 83% about follow-up treatment and 52% correctly knew what examinations were planned. Accurate knowledge about medication is much lower. Whereas 51% of the patients who received medication knew the purpose of the medication, only 37% recalled the name, 35% correctly knew how to take the medication, 30% held accurate information about side effects and 21% knew how long to take the medication.
Comprehensiveness index for the discharge conversation
The overall comprehensiveness of information provided by the physician during the discharge conversation was evaluated using the index developed by the ED experts. We assigned one point for each of the following topics discussed: (a) diagnosis, (b) examinations performed, (c) impending examinations and (d) follow-up treatment. The results in table 2 show sufficient or good information comprehensiveness (three or four points) in 83%, and insufficient or very insufficient completeness (one or two points) in 17% of the cases.
Informedness index of patients after discharge
Table 3 shows the proportion of patients who knew 0, 1, 2 or 3 components of the information according to the patient informedness index, independent of the source of the information (discharge conversation or other sources). The index includes: (a) information about diagnosis, disorder or affected organ, (b) examinations planned and (c) further treatment. None of the patients left without any information about the three main domains. However, 18% of the patients had only one information, 39% had two and less than half of the patients left with complete information (43%).
Mean patient satisfaction with the discharge conversation was 4.7 (scale from 1 to 5, SD=0.64). The Spearman Rho correlation showed no significant relationship between patient satisfaction and the comprehensiveness index (R=0.06; p=0.59) or between patient satisfaction and the informedness index of the patient (R=0.05; p=0.66).
Mean satisfaction score of the physicians with the discharge conversation was 4.24 (scale from 1 to 5; SD=0.80). The Spearman Rho correlation between physician satisfaction and comprehensiveness index was not significant (R=0.07; p=0.52); as the comprehensiveness index refers to the information given by the physicians, this lack of association indicates that the physicians were not aware of any lack of information transmitted.
Communication and information transmission are central elements in the physician–patient relationship,1 and this is particularly important for emergency medicine, where patients are often insecure and depend on reassurance by quantitatively adequate and qualitatively adapted information. We used an unstratified patient sample from the ED to evaluate whether the information provided in the discharge conversation met this need.
We were particularly interested in how much information was provided by the physician. Important aspects such as details of the disease, disorder or injury and follow-up treatment were discussed with most patients (90%–99%). However, performed and planned examinations were discussed with only about half of the patients, and the amount of information given concerning medications was clearly insufficient. These results are similar to other studies.16 According to the expert rating there was insufficient information transfer in 17% of the discharge conversations. These results contrast with the finding that the physicians were generally satisfied with their discharge conversations.
Reasons for the lack of communication of medical information may lie in the fact that the focus of emergency physicians is primarily on rapid and correct diagnosis, and on a continuous flow of cases. Time pressure could also play a role. However, in this study, physicians spent an average of 4 min on the discharge conversation, which is about three times as much as the 76 s duration found in another study in the emergency sector.16
Another important finding of this study is that only between 30% and 82% of the information discussed during the discharge conversation was correctly recalled by patients immediately after discharge. This loss of information was only partially compensated by other sources. According to the informedness index, which includes information acquired from discharge conversation but also during treatment, 66% of patients lacked information about one of the main domains (diagnosis, examinations planned and follow-up treatment) at discharge (not accounting for detailed knowledge about medication). These patients therefore left the ED with insufficient information about their health problems and further treatment. Two-thirds of patients with information deficit is clearly too high. However, this does not seem to be specific to our ED, as it corresponds to the results of a comparable study by Engel and colleagues, who found information deficits at discharge for over 70% of patients.3
Similar to other studies, patient satisfaction with the discharge conversation was high. This could be related to the bias towards generally positive answers, due to the patient's dependence on the medical personnel,17 and it could be related to nice and respectful interpersonal treatment. However, the high satisfaction could also reflect earlier findings that patients are not aware of information deficits;3 this is corroborated by the absence of a statistically significant correlation between objective information transmission and patient satisfaction with the information transmitted. For these reasons, patient satisfaction or surveys should not be used as the sole indicator for the quality of treatment and care.18 Rather, quality and effectiveness of medical treatment and care should be assessed through observational data, as we did in this study.
This study has limitations. First, patients with communication restrictions due to medical or language reasons did not participate in the study. These, however, are precisely the patients who often come from other cultures for whom thorough information would be especially important. Second, approximately a fifth of eligible patients refused participation, which can lead to a non-response bias that must be taken into account in the interpretation of the results.19 Third, physicians were aware that the discharge conversation was recorded; they might thus have paid particular attention to the discharge, which might induce a positive bias.
Despite these limitations, conclusions about ways of improving information transmission can be drawn. One the one hand, further training of physicians to overcome communication deficits is required. Physicians in EDs must be aware of the limited memory span of stressed patients in an emergency situation, and they must be aware that information provided during treatment is often not sufficient to inform patients. Thus, providing comprehensive information during the discharge conversation might be worthwhile, even if it implies repeating previously discussed aspects. On the other hand, the quality of discharge conversation could be improved by establishing clearer rules. A standardised procedure with a checklist, and more written discharge information, would contribute to better patient informedness.
Competing interests None.
Patient consent Obtained.
Ethics approval The ethics approval was obtained by University Hospital of Bern, Switzerland.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Data are available on request from the coauthor: franziska tschan firstname.lastname@example.org