[Hospital readmission in internal medicine]

The readmission rate could be a valuable tool as measurement of hospital quality. Readmissions are due to several factors: clinical, hospital related and patient related. We analyze readmission to internal medicine in a hospital of third level.

During 11 months in 1988 we counted all readmissions (R) defined as every previous admission occurred in a span of five years into an area of internal medicine composed by 8 beds. We counted number of readmssions, time from the last readmission, living area (city vs country), sort of primary care physician (GP vs family care specialist), living way (single, with family, institution, homeless). Precipitating factors were observed as well as diseases causing it. R were classified as R related (RR) when readmission was provoked by the same pathological condition or a complication. Multi-readmission (MR), those R caused by the same disease process and treated in different areas and services of the hospital. Avoidable R (AR), those R which did not fullfil AEP criteria. Early readmission (ER) those R occurring before 30 days after last discharge.

Three hundred and eleven patients (mean age 67.93 (SD 15.51), males 64%, mean length of stay 7.75 (SD 4.35), 93% admitted from emergency yard, mortality rate 3.5%) were included. R were 111 (35.5%), RR 83 (26 and 75% of RR), MR 68 (82% of RR), ER 33 (39.7% of RR) and AR 16 (19.2% of RR) patients. The most frequent diseases were heart failure and chronic respiratory diseases. Main causes of R were worsening of chronic disease 41 (37%), non-appropriale ambulatory management 24 (22%) erroneous diagnosis 8 (7%), iatrogenic effect 7 (6%), new disease 29 (26%) and others 2 (2%). Mortality rate in R patients was 7.2% (confidence interval 95% 2 to 9%). Number of readmissions were 3.22 (SD 2.25) and time to readmission 8.99 (SD 11.96) months. Living in city (p < 0.05) and to be cared by family physician (p < 0.01) both were factors accelerating readmission. Patients with RR had a higher number of readmissions (3.55 SD 2.23 p < 0.001) and they occurred sooner (8.03 SD 11.85) (p < 0.01). There was a trend to higher readmission rate in female (p 0.052). Fifty-seven percent of RR patients did not have consultation with primary care physician (p < 0.05) (confidence interval 95% 3 to 39%). Consultation with primary care yielded a delay in readmission of 5 months (p < 0.01). Patients with MR had an increased number of readmissions (p < 0.01). Associated factors were iatrogenic effect (p < 0.05), non-appropriate ambulatory management (p < 0.001) and worsening chronic disease (p < 0.001). Patients with ER were readmitted 0.45 (SD 0.30) months after the last discharge and they had a higher mortality rate (p < 0.05). Patients with AR had a mean length of stay shorter (p < 0.05), a trend to higher readmission rate (p = 0.06) and sooner (p = 0.08) with a null mortality rate (p < 0.01). As risk factors for RR in logistic regression were identified MR, AR, ER and causes of readmission consisting in worsening of chronic disease, non-appropriate ambulatory management, erroneous diagnosis and iatrogenic effect.

Our readmission rate is 26%, chronic respiratory diseases and heart failure being the main diseases. Over 39% of causes of readmission could be preventable and there is a facilitation phenomenon in number and time to readmission caused by previous readmissions. Risk factors for readmission in internal medicine are multi-readmission, early and avoidable readmission and as specific causes worsening of chronic disease, non-appropriate ambulatory management, erroneous diagnosis and iatrogenic effect.