Background:
Substantial efforts are being made to reduce readmissions. However, previous studies of causes for readmissions have not solicited input from physicians involved with patient care both in the hospital and in primary care settings.
Methods:
This was an observational study of patients readmitted within 30 days to general medical services at 13 academic medical centers. At the time of readmission, we e‐mailed or faxed surveys to each patient's primary care physician, the previous inpatient physician, and current attending in order to solicit their impressions of potential causes of the readmission and areas for improvement in care. Each physician was e‐mailed or faxed a survey up to 5 times within 14 days of the readmission; we encouraged physicians to review medical records and speak with their teams to complete each survey. We used univariable statistics to characterize overall reports of potential causes and kappas to assess agreement between physicians in individual cases.
Results:
Interim analysis point — patient inability to manage symptoms was highly rated by all MD groups, with similar proportions of MDs citing this potential cause. However, other causes were rated differently between MD groups. For example, PCPs rated communication between MDs, timeliness of discharge summaries, and inappropriately rapid discharge as potential causes more often than inpatient MDs. Readmitting MDs were more likely to feel that inadequate or incomplete treatment at the initial admission and inadequate home services were potential causes. Discharging MDs more often attributed readmissions to inappropriate decisions in the ED or failure of patients to keep follow‐up appointments. MD agreement within individual patients as estimated by the kappa statistic was generally poor.
Conclusions:
Inpatient‐based and primary care physicians generally do not agree on causes of readmission in individual cases. However, aggregated “360‐degree” data allow for learning from all different perspectives and may help to frame effective transitions‐of‐care interventions. Assessments of causes and preventability of readmissions should include a diverse range of perspectives in order to reduce bias.
PCP, Discharging Physician, and Admitting Physician Reports of Potential Causes of Readmission
| PCP | Discharging MD | Readmitting MD | n (%) Attributing the factor as a potential contributing cause for readmission (overall n = 130) |
| Patient understanding and ability to manage care | |||
| Patient or caregiver inability to manage his/her symptoms | 45 (34.6%) | 49 (37.7%) | 48 (36.9%) |
| Patient or caregiver inability to otherwise care for him/herself | 29 (22.3%) | 22 (16.9%) | 21 (16.2%) |
| Patient or caregiver lack of understanding of the postdischarge plan | 14 (10.8%) | 11 (8.5%) | 11 (8.5%) |
| Patient or caregiver inability to manage his/her medications | 18 (13.9%) | 22 (16.9%) | 12 (9.2%) |
| Insufficient or ineffective patient or caregiver education | 9 (6.9%) | 5 (3.9%) | 6 (4.6%) |
| Continuity of care and communication | — | — | — |
| Insufficient monitoring of the patient's condition(s) after discharge | 20 (15.4%) | 11 (8.5%) | 19 (14.6%) |
| Insufficient communication with post–acute care provider(s) re postdischarge plan | 15 (11.5%) | 2 (1.5%) | 2 (1.5%) |
| Failure to obtain an appropriately timely follow-up appointment or follow-up studies | 15 (11.5%) | 10 (7.7%) | 7 (5.4%) |
| Discharge summary poorly written or with missing or erroneous information | 4 (3.1%) | 1 (0.8%) | 0 (0%) |
| Inability of the patient to keep the follow-up appointment or follow-up studies | 6 (4.6%) | 12 (9.2%) | 10 (7.7%) |
| Social supports | — | — | — |
| Inadequate support for nonclinical issues (such as food, heat, transportation, or inability to afford medications) | 16 (12.3%) | 11 (8.5%) | 10 (7.7%) |
| Inadequate home services or equipment after discharge | 9 (6.9%) | 5 (3.9%) | 13 (10.0%) |
| Inappropriate choice of discharge destination | 7 (5.4%) | 7 (5.4%) | 4 (3.1%) |
| Problems with care | — | — | — |
| Inappropriate/inadequate treatment of the patient during the index admission | 9 (6.9%) | 5 (3.9%) | 12 (9.2%) |
| Misdiagnosis made during the index admission | 3 (2.3%) | 3 (2.3%) | 6 (4.6%) |
| No or inadequate end-of-life or goals of care planning | 5 (3.9%) | 3 (2.3%) | 4 (3.1%) |
| PCP, discharging physician, and admitting physician reports of potential contributing causes to readmission | 4 (3.1%) | 9 (6.9%) | 5 (3.9%) |