Background: Hospital discharge instructions, written by healthcare providers for the after visit summary (AVS), provide patients with essential information for post-hospital care. However, AVSs are often hindered by language barriers, medical jargon, inaccurate medication lists and follow up instructions, and a lack of disease-specific and home care instructions. We reviewed discharge instructions from one academic medical center to assess quality and adherence to known best practices.
Methods: We reviewed discharge instructions from patients discharged over one month from one institution’s inpatient attending-only and resident Medicine services. Two reviewers independently assessed readability by evaluating for language above the 8th grade reading level. Using a structured checklist, the utilization of an institution-specific template, presence of disease-specific and home health/ancillary care instructions, consistency of written medication lists (compared with auto-populated lists elsewhere in the AVS), completion of outpatient follow-up instructions, and completeness of translated instructions were quantified. Finally, semi-structured interviews with 9 healthcare workers (4 nurses and 5 physicians) were performed to identify areas for improvement.
Results: Discharge instructions from 81 adult inpatient encounters were reviewed, of which 41 (51%) patients were discharged from attending-only services and 49% from resident teams. Most (94%) discharge instructions used a standardized template with the following sections: reason for admission, hospital course, discharge medications, follow-up care, and return precautions. Nearly a third (27%) of discharge instructions included medical jargon without explanation or were written above an 8th grade reading level, and only 57% stated the discharge diagnosis in patient friendly language (Table 1). Only 21% of discharge instructions included disease-specific care instructions, though 43% had attached supplementary handouts. Only 9% of discharge instructions included home health contact information. Over a third (37%) of instructions contained medication inconsistencies when compared to auto-populated lists elsewhere in the AVS. While 51% of AVSs mentioned outpatient referrals, only 26% contained appointment information. Of the 13 AVSs that included translated instructions, 31% had incomplete translations or translation errors. Semi-structured interviews with nurses revealed themes of confusion around medication changes, disease-specific care, and follow-up instructions. Interviews with physicians also revealed frustration regarding unclear follow-up and referral information.
Conclusions: Discharge instructions at one institution fell short of best practices across several key domains, despite high utilization of a standardized template. Future work can aim to improve (1) medication documentation consistency; (2) patient-friendly language and descriptions; (3) clear follow-up and home health information; (4) completeness of language concordance across populations; and ideally, (5) language tailored to a patient’s specific health literacy level. Given its high utilization, improving the existing institutional template may represent an opportunity to achieve these changes. Alternatively, leveraging artificial intelligence platforms to compose discharge instructions may present an opportunity for improvement across many of the stated domains while also minimizing physician burden.
.jpg)