We describe a systematic approach to the identification and classification of near-miss events on labor and delivery in a large, national health care system. Voluntary reports of near-miss events were prospectively collected during 2010 in 203,708 deliveries. These reports were analyzed according to frequency and potential severity. Near-miss events were reported in 0.69% of deliveries. Medication and patient identification errors were the most common near-miss events. However, existing barriers were found to be highly effective in preventing such errors from reaching the patient. Errors with the greatest potential for causing harm involved physician response and decision making. Fewer and less effective existing barriers between these errors and potential patient harm were identified. Use of a comprehensive system for identification of near-miss events on labor and delivery units have proven useful in allowing us to focus patient safety efforts on areas of greatest need.
Avoidance of medical errors is a key component of quality medical care. However, because the complete elimination of medical error is not achievable, effective patient safety programs seek to construct barriers that prevent errors from reaching the patient and causing harm. The creation of such barriers is facilitated by the identification of near-misses or close calls, defined as unplanned events caused by error that do not result in patient injury but have the potential to do so. Such near-miss events are, in most systems, many times more common than injury events, yet are generally much more difficult to identify and quantify.
We report our experience with a system of near-miss identification and analysis on labor and delivery units in the largest obstetric health care delivery system in the United States. This system has allowed us to identify and prioritize efforts to reduce patient harm, and represents the first reported comprehensive analysis of near-miss events in obstetrics.
Materials and methods
The Hospital Corporation of America (HCA) has an annual delivery volume of over 200,000 deliveries in 104 hospitals in 21 states. Such facilities range from small, rural hospitals to major academic medical centers and are reasonably representative of the overall US health care system. In 2008, HCA developed and deployed an electronic voluntary reporting system aggregating deidentified close call and event reports into a central repository called the Patient Safety Improvement Program (PSIP) database that is among the largest event repositories in the United States. Using a standard taxonomy derived from existing external and internal programs and narrative reports, this reporting system became a resource for collective learning and identification of high-risk processes that have the potential to cause patient harm. This program encourages all providers to report both unplanned events, (defined as any unplanned occurrence, which is undesirable, untoward, or unanticipated) and near-misses, (defined as unplanned events that have the potential to cause harm, but do not do so). The clear overlap in these definitions is purposeful, and directed at the identification of all events that could, under closer scrutiny, possibly represent actual or potential harm to a patient. Although this is a nonpunitive system that allows the reporting of near-miss events without fear of individual sanction or disciplinary measures, every effort is made to address contributing factors related to human error and system flaws identified through in-depth report analysis.
After collection of all reported events from labor and delivery units during 2010, a random sample was selected for analysis using Cochran’s sample size formula set for a 5% maximum error. All reports were initially reviewed and preliminarily categorized as to the potential for patient harm by a team consisting of nursing and administrative personnel. Each report was then additionally reviewed and analyzed by the principle physician author. The selected reports were divided into 3 groups: unplanned events not related to medical error, medical errors resulting in temporary or permanent patient harm, and near-misses, as defined above. The near-miss group was then subjected to detailed analysis and is the subject of this report.
Each near-miss event was categorized as to absolute and relative frequencies as outlined in the Table . Events were then given a score for each of several defined parameters in a modification of a previously described system known as failure modes and effects analysis. In each case, a low score is desirable and a high score undesirable. These parameters were defined as follows:
- A
Worst possible outcome. Each event was given a score from 1 (negligible chance of harm had the event reached the patient) to 5 (high likelihood of serious harm had the event reached the patient.) An example of the former would be a charting error indicating that 1 g of cefazolin had been given for cesarean prophylaxis when in fact 2 g had been given. An example of the latter would be the ordering of penicillin in a patient with a history of penicillin anaphylaxis, which was caught by standing nursing and pharmacy protocols mandating referral to allergy history before administration of any drug.
- B
Method of detection. Each event was given a score of 1 (error detected by existing monitoring systems) or 2 (error detected fortuitously).
- C
Number of barriers in place. Each event was given a score of 1 (single barrier in place aimed at preventing error from causing patient harm) or 2 (more than 1 barrier in place preventing error from causing harm).
- D
Quality of barriers in place. Each event was given a score of 1 (high quality barrier[s]) in place or 2 (low quality barrier[s]) in place. This was based on our knowledge of our internal error-prevention systems, as well as our observation of reported actual harm events for the various categories of errors in the broader database.
| Error type | No. (%) | Hazard score |
|---|---|---|
| Medication error | 120 (33) | 3 |
| Patient identification | 95 (19) | 2 |
| Failure/delay in obtaining laboratory | 57 (11) | 6 |
| Failure to respond | 49 (10) | 15 |
| Failure to follow policy/protocol | 46 (9) | 9 |
| Charting error | 31 (6) | 2 |
| Equipment failure | 28 (6) | 5 |
| Slip/fall | 27 (5) | 11 |
| Information transfer | 23 (5) | 10 |
| Physician performance error | 14 (3) | 17 |
| Incorrect test ordered | 7 (1) | 1 |
| Inadequate staffing | 4 (<1) | 12 |
| Error in test interpretation | 2 (<1) | 3 |
| Wrong procedure ordered | 3 (<1) | 4 |
| Laboratory error | 3 (<1) | 1 |
| Retained foreign body | 3 (<1) | 5 |
After the initial scoring of each event, a mean score for events in each category was calculated (termed A, B, C, and D) and used to create a new parameter for each category of event, the hazard score (A × B × C × D) ( Table ) (individual scores for each near-miss event category available on request).
This analysis included only deidentified data examined for purposes of quality improvement and was thus exempt from institutional review board review based on 45CFR46.101(b)2 and 46.102(f) and 45CFR164.514(a)-(c) of the Health Insurance Portability and Accountability Act.
Results
During 2010, 3959 unplanned events were reported on labor and delivery of 203,708 total deliveries (1.9%). More than 95% of reports originated from nursing or laboratory staff. Using the methodology described above, 1248 events were randomly selected for analysis. Seven hundred eighty-five events either did not involve medical error (for example, precipitous delivery not attended by a physician or certified nurse midwife within a few minutes of patient arrival in the hospital) or was a medical error resulting in patient harm (for example, fetal scalp injury resulting from misuse of forceps). Four hundred sixty-three reports involved near-miss events as defined above, and were analyzed in detail. When extrapolated to the entire delivery population during 2010, near-miss events were reported in 0.69% of deliveries.
The Table demonstrates the types and frequency of near-miss events on labor and delivery as well as the composite hazard score for each type of event.
Results
During 2010, 3959 unplanned events were reported on labor and delivery of 203,708 total deliveries (1.9%). More than 95% of reports originated from nursing or laboratory staff. Using the methodology described above, 1248 events were randomly selected for analysis. Seven hundred eighty-five events either did not involve medical error (for example, precipitous delivery not attended by a physician or certified nurse midwife within a few minutes of patient arrival in the hospital) or was a medical error resulting in patient harm (for example, fetal scalp injury resulting from misuse of forceps). Four hundred sixty-three reports involved near-miss events as defined above, and were analyzed in detail. When extrapolated to the entire delivery population during 2010, near-miss events were reported in 0.69% of deliveries.
The Table demonstrates the types and frequency of near-miss events on labor and delivery as well as the composite hazard score for each type of event.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
