08

Oct 2024

A study indicates that harmful diagnostic errors could affect 1 in every 14 patients in general medical hospitals.

Published in General on October 08, 2024

Harmful diagnostic errors could be affecting up to 1 in every 14 hospital patients—specifically those receiving general medical care—according to findings from a single-center study in the US published online in the journal BMJ Quality & Safety.

According to the researchers, most of these errors (85%) are likely preventable, highlighting the urgent need for new strategies to enhance surveillance and prevent such mistakes from occurring.

Earlier reports suggest that current trigger tools used to identify medical mistakes are insufficient for detecting harmful diagnostic errors, particularly those with less severe outcomes. In response, the researchers developed and validated a structured case review process that enables clinicians to analyze the electronic health record (EHR) to evaluate the diagnostic process for hospital patients, assess the likelihood of a diagnostic error, and determine the impact and severity of the harm caused.

They applied this process to retrospectively estimate the prevalence of harmful diagnostic errors in a randomly selected sample of 675 hospital patients from a total of 9,147 who received general medical care between July 2019 and September 2021, excluding the peak of the COVID-19 pandemic (April–December 2020).

Cases identified as high-risk for diagnostic errors included transfers to intensive care 24 or more hours after admission (130 cases; 100%); deaths within 90 days of admission, whether in the hospital or after discharge (141 cases; 38.5%); and complex clinical issues without transfer to intensive care or death within 90 days (298 cases; 7%).

Complex clinical issues encompassed clinical deterioration, treatment by multiple medical teams, unexpected events such as cancelled surgeries, and unclear or inconsistent diagnostic information recorded in the medical notes.

The 106 cases classified as low risk (2.5%) did not meet any of the high-risk criteria.

Each case underwent review by two trained adjudicators who evaluated the likelihood of a diagnostic error and identified associated process failures using the Diagnostic Error Evaluation and Research Taxonomy modified for acute care. Harm was categorized as minor, moderate, severe, or fatal, and the assessors also evaluated whether the diagnostic error contributed to the harm and if it was preventable. Cases with discrepancies or uncertainty regarding the diagnostic error or its impact were subjected to further review by an expert panel.

Among all reviewed cases, diagnostic errors were identified in 160 instances involving 154 patients. These errors included intensive care transfers (54), deaths within 90 days (34), complex clinical issues (52), and low-risk patients (20).

Harmful diagnostic errors were determined to have occurred in 84 cases (82 patients), broken down as follows: 37 cases (28.5%) among intensive care transfers, 18 cases (13%) among patients who died within 90 days, 23 cases (8%) among those with complex clinical issues, and six cases (6%) in low-risk patients.

The severity of harm was classified as minor in five cases (6%), moderate in 36 cases (43%), major in 25 cases (30%), and fatal in 18 cases (21.5%). Overall, it was estimated that 85% of harmful diagnostic errors were preventable, with older, white, non-Hispanic, non-privately insured, and high-risk patients identified as most vulnerable.

When weighted to reflect the population, researchers estimated that the prevalence of harmful, preventable, and severely harmful diagnostic errors in general medical hospital patients was just over 7%, 6%, and 1%, respectively.

Process failures were significantly linked to diagnostic errors, particularly uncertainties in initial assessments and complex diagnostic testing and interpretation (four times the risk), suboptimal subspecialty consultations (three times the risk), patient-reported concerns (three times the risk), and deficiencies in history taking (2.5 times the risk).

Among the errors, 40 cases (48%) related to the primary diagnosis at admission or discharge, and 44 cases (52.5%) to a secondary diagnosis; 52 cases (62%) were classified as delays. Errors resulting in major or fatal harm were common in the high-risk group (55%, 43 out of 78) but rare in the low-risk group (0 out of 6).

The most frequently associated diagnoses with these diagnostic errors included heart failure, acute kidney failure, sepsis, pneumonia, respiratory failure, altered mental state, abdominal pain, and hypoxemia (low blood oxygen levels).

The researchers suggest that careful error analysis and the integration of AI tools into clinical workflows could help reduce their prevalence by enhancing monitoring and enabling timely interventions.

This is an observational study based on estimates derived from data on patients receiving general medical care at a single centre, and the researchers advise that findings should be interpreted within this context. They also acknowledge that the sample was limited to patients with hospital stays under 21 days and that the study relied on information captured in electronic health records, which may inaccurately record deaths within 90 days.

Moreover, for patients and their families looking for accommodation, considering hospital stay accommodation options can significantly alleviate stress during this challenging time. Various facilities in Melbourne offer supportive environments and comfortable stays that cater to the needs of patients and their relatives.

Nonetheless, they conclude, "We estimate that a harmful diagnostic error occurred in one of every 14 patients hospitalized in general medicine, the majority of which were preventable. Our findings underscore the need for novel adverse diagnostic error surveillance approaches."