The History of 34 Errors Identified in 874 Patients Analyzed at Weekly Clinical-Pathological Meetings in Two Institutions over 22 Years (1982–2004).

Clinical-pathological sessions are a good method for solving diagnostic and/or therapeutic problems in patients with hemopathies. In these sessions, errors made during health care given to the patient can be detected. By analyzing how and why these errors are made, we can improve patient care and prevent further mistakes. The objective of this study is to describe 34 errors identified in 874 patients with hemopathies at clinical-pathological meetings in two centers, performed with the aim of solving a diagnostic problem and/or a therapeutic decision or due to a great interest of the case, in two institutions, over 22 years (1982–2004). An intererdisciplinary team of hematology specialists gathered every week at interactive sessions of about 45 minutes each, in both institutions. The methodology of sessions was: a description of the medical history of a patient in a one or two-page report and a revision of the different samples (peripheral blood, bone marrow and lymph node morphology, immunocytochemistry, flow-cytometry, cytogenetic and molecular studies) with the aid of a microscope and a TV monitor. A diagnostic and/or treatment were proposed at the end of the session. Eight-hundred and seventy-four reports were analyzed. All the diagnostics were classified: chronic lymphoproliferatives disorders (445), myeloproliferative and myelodysplastic syndromes (136), acute leukemias (136), other haematological diseases (74), non-haematological diseases (31), without a diagnosis after the meeting (52). We identified diagnostic (D) and therapeutic (T) mistakes and considered as the main causes of the medical error (mistake in the diagnosis and/or treatment): lack of expertise (LE), malpractice (MP), impetuosity (IM), bad logistic support (LS), inexplicable (IN). We divided the 22 years into two decades and each error was classified in one of these two groups. Our own mistakes (OM) and the errors made in other institutions (OI) were identified. A comparison between number of errors made in the first 11 years and the second 11 years was made using a Chi-square test. P<0.05 was considered statistically significant. Thirty-four errors (4 %) were detected, being more D and T (20) rather than only D (14). The type of error detected was: 17 LE, 7 MP, 5 IM, 3 LS, 2 IN. Twenty errors were OM and 14 were made in OI. The difference in the proportion of errors detected during the first and the second decade (6.6 % vs 2.8 %, respectively) was statistically significant (p=0.05). No error led to the death of any patient or were life-threatening in any way. Errors may be made in the diagnosis and treatment of hematologic patients. Although the rate of error found appears to be high, it can be considered as low, since the cases were presented in scientific sessions because of diagnostic and/or therapeutic problems. One of the best ways of improving how to care for future patients is to detect and analyze the errors made. Many lessons can be learnt in this way.