[Adverse events and preventable consequences: retrospective study in five large Italian hospitals]

To define the incidence of adverse events and their preventability in a representative sample of patients in five acute hospitals located in the North, the Centre and the South of Italy. Other objectives include the evaluation of the consequences of adverse events and their distribution according to specialties.

Retrospective and multicentre study. The methodology is focused on the review of clinical records related to hospital admissions in the year 2008 with a sample of 5 hospitals belonging to the national healthcare system selected according to criteria of location (North, Centre and South of Italy) and complexity (regional reference hospitals). The clinical records included in the study were selected in a random way starting from the electronic archives of the hospital discharges of each participating centre.

7,573 clinical records were reviewed with a process of two stages managed by two reviewers each. The first stage of the review process involved 7 physicians, 1 nurse, 1 pharmacist and 1 biologist with skills and experiences in clinical risk management and in analysis of clinical documentation. The second stage was realized by 10 physicians (5 specialists, 3 experts in public health and 2 forensic physicians), also for the second stage every person involved had specific training in clinical risk management. The reviewers attended a 20-hour training course.

Study of the incidence of adverse events identified during the admissions included. In the case of more than one adverse event for each admission, it is calculated the cumulative incidence of adverse events for each patient. We also considered the percentage of re-admitted patients for each adverse event, the percentage of adverse events which occurred in the phase of pre hospitalization and the degree of preventability of adverse events. A description of the identified adverse events was realized. The sample of the data included in the study was described in terms of included and excluded subjects with respect to the planned research design. Different products and results were tested and validated in the study and could be reused in the future research products.

The overall average of the incidences of adverse events was 5.2%, the median was 5.5% and it is consistent with the expected results mentioned in the protocol of the study. The identified incidence of adverse events is lower than the median rate of international studies (9.2%). The distribution of adverse events for specialties underlines the majority of adverse events in the medical area (37.5%), in opposition to the results of other studies; the surgery is the second specialty for number of adverse events (30.1%), followed by the emergency room (6.2%) and obstetrics (4.4%). The study identified 56.7% of adverse events as preventable. The consequences of adverse events were classified in different typologies: the prolonged stay was the most frequent consequence, followed by the disability at discharge. The death of the patient had a median occurrence of 9.45%. The concordance between the two reviewers in the evaluation of the clinical records was very high (higher than 95%) except for two centres.

The results were consistent with the results of other international studies similar in scope in terms of type of study (definition of the rate of adverse events) and epidemiological study design (retrospective study). The incidence rate, previously mentioned as 5.2%, coincides with the unfavorable rates of events determined in varied countries. The preventability resulted in an average of 56.7%. The variability of the results obtained in our inquiry are likely attributable to varied factors occurred during the study.