Abstract
Objective:
The aim of this study was to identify factors associated with the occurrence of adverse events (AE) or medical errors (ME) during inpatient psychiatric hospitalizations.
Methods:
A full-probability random sample of 4,371 charts from 14 inpatient psychiatric units at acute care general hospitals in Pennsylvania were reviewed in a two-stage process that comprised screening and flagging by nurses followed by review by psychiatrists. AE and ME rates were calculated overall and then stratified by patient and hospital factors. Unadjusted and adjusted logistic regression models examined predictors of AEs and MEs.
Results:
An AE was identified in 14.5% of hospitalizations (95%CI=11.7–17.9) and an ME was identified in 9.0% (95%CI=7.5–11.0). In adjusted analyses, patients with a longer length of stay and older patients had higher odds of experiencing an AE or an ME. Patients ages 31–42 (compared with ages 18–30), with commercial insurance (compared with Medicare or Medicaid or uninsured), or treated at high-volume hospitals (compared with low, medium, or very high) had lower odds of an AE. Patients age 54 or older (compared with ages 18–30), admitted during the weekend, admitted to rural hospitals (compared with urban), or treated at very-high-volume hospitals (compared with high) were more likely to experience an ME.
Conclusions:
This study provides insight into factors that put patients and hospitals at increased risk of patient safety events. This information can be used to tailor improvement strategies that enhance the safety of patients treated on general hospital psychiatric units.
INTRODUCTION
More than 17 years have passed since the publication of the groundbreaking patient safety report ‘To Err Is Human’(1), which called international attention to the issue of adverse events (AE) and medical errors (ME) that occur in the care of hospitalized patients. AE are defined as the negative unintended consequences of clinical care that lead to injury, impairment, or other harm (2, 3). ME are defined as the commission or an omission of clinical care with potentially negative consequences for a patient that would have been judged wrong by skilled and knowledgeable peers at the time it occurred, regardless of whether it caused harm (4). Since that time, patient safety research has guided the development of important interventions to prevent ME and lower the incidence of AE in general medical and surgical care (5–14). Knowledge about these critical patient-safety events, however, is notably scarce for mental health care because major studies on the topic have systematically excluded patients receiving psychiatric care in acute care general hospitals. This lack of information about inpatient mental health care has hindered the extension and adaptation of patient safety lessons learned in general medicine.
There has been some research on the incidence of specific types of AE and ME in mental health care. For example, patient suicide and falls in inpatient psychiatry have been studied (15, 16). However, these events are often studied in small samples, without a tested methodology, and in isolation (that is, absent a broader spectrum of inpatient psychiatric patient safety events). One large, national study systematically examined the incidence, nature and preventability of patient safety events in inpatient psychiatric hospital settings. That study of more than 8,000 discharged patients found that one of five patients receiving care on an inpatient psychiatric unit experienced an AE or ME and that 56.6% of all events were characterized as preventable (17). Although the study provided important data about patient safety rates, generalizability is limited because it was conducted only at Veterans Health Administration (VHA) hospitals which deliver care to a very specialized segment of the population and it did not examine any of the patient and hospital factors associated with a higher risk of patient safety events. Identifying potential risk factors could inform targeted efforts to reduce the incidence of these events.
Community-based acute care general hospitals are the primary service system for inpatient psychiatric care. Each year there are more than one million discharges from inpatient psychiatric units in the United States, and approximately half of these discharges are from acute care nonfederal general hospitals (18). The other half are from state mental hospitals, freestanding psychiatric hospitals, and VHA hospitals – all of which represent different patient populations and systems of care. Given that acute care general hospitals are the most common providers of inpatient psychiatric care, the wide variability in the structural characteristics of general hospital settings, as well as the demographic and clinical characteristics of the patients they care for, studying AE and ME in this heterogeneous clinical care environment is integral to understanding the broad context in which they occur. The study reported here examined MEs and AEs in inpatient psychiatric settings at general hospitals in a large, diverse sample of hospitals in Pennsylvania, as well as the patient and hospital factors that influenced their occurrence.
METHODS
Study sample
This study included medical records of patients discharged from psychiatric inpatient units at 14 acute care general hospitals in Pennsylvania during 2010. The Pennsylvania Health Care Cost Containment Council (PHC4) provided a dataset from which a stratified random sample of inpatients from psychiatric units in general hospital settings was drawn. The PHC4 data included detailed information about patient demographic and clinical characteristics (e.g. diagnosis and procedure codes, length of stay and hospital information) for all hospital discharges in the state. From these data, a random sample of 19 general hospitals was selected stratified by teaching status and hospital location (urban/rural) with probabilities proportional to each hospital’s number of discharges. Next, a subsample of approximately 300 patient discharges was drawn from each selected hospital with probabilities inversely proportional to the size of the hospital. Long-term admissions with a length of stay of more than 90 days were excluded from the sample. Because five hospitals declined to participate, our total sample included 14 inpatient psychiatric units, representing a 74% recruitment rate. There were no significant differences between the included and excluded hospitals in teaching status, urban/rural location, or size. Discharges from the responding hospitals were weighted to account for non-response and to be representative of all discharges from psychiatric units at Pennsylvania acute care general hospitals.
Design and procedure
We conducted a retrospective cross-sectional chart review with two stages: nurses trained as medical record administrators (MRAs) screened and “flagged” the random sample of medical records within each hospital for potential patient safety events, and trained physician reviewers then evaluated “flagged” charts for harm, error, and preventability. We developed training manuals and instrumentation to implement both tiers of the study on the basis of the methodology used in the landmark patient safety studies (2, 19). We then recruited 11 MRAs across the state to review medical records from each study hospital and flag them as described above. After extensive training, the teams of MRAs reviewed 4,371 of the 4,401 medical records (99.3%); several charts were unavailable or unable to be located. Each medical record was carefully evaluated by using information in admission and discharge notes, clinical notes, nursing notes, progress notes, physician orders, and medication administration records. We recruited and implemented our rigorous training process with 12 physicians across the state. Throughout the study, reviewers examined an overlap of 10% of the sampled charts and these were used as training files during weekly calls to maintain reliability. Our study team of reviewers abstracted data from the medical records of all hospitals in the sample, with the exception of one hospital that stipulated that only its employees could have access to the data. For that hospital we trained a separate team of nurse and physician reviewers in the study methods. None of the reviewers had treated the patients whose records they reviewed.
Outcome measures
Patient safety events were categorized as dichotomous outcomes for the occurrence of adverse events (AE) or medical errors (ME). The following events were categorized as AE: adverse drug event (ADE), self-harm or injury, assault, sexual contact, patient fall and other. MEs included: any mention in the chart of medication errors, elopement, possession of contraband and other non-medication errors. Detailed definitions of these events are included in an online supplement to this article, and the events have been described elsewhere (17).
Predictors
For each discharge, we examined patient demographic and clinical factors, as well as characteristics of the hospital from which the patient was discharged. We investigated data on gender, age (18–30, 31–42, 43–53, 54+) and race (white, non-white). Clinical factors included admission day (weekday vs weekend), length of stay (1–3, 4–6, 7–9, 10+ days), insurance status (uninsured, commercial or Medicaid/Medicare). The ICD-9 coding system was used to categorize principal diagnosis, which included psychosis (295, 297, 298), mood disorders (296) and other diagnoses (all other codes). Suicidal ideation (V62.84) or suicide attempts (E950-E959) were identified using all available diagnosis codes on the discharge claim. Drug use was categorized as continuous (303.01, 303.91 304.21–304-.91, 305.01), unspecified or episodic (303.00, 303.90, 304.20–304.90, 305.20, and 303.02, 303.92, 304.22–304.92, 305.02), or none/in remission (303.03 303.93, 304.23–304.93, 305.03). Hospital-level factors included teaching status (teaching, non-teaching), urbanity (urban or rural), and low, medium, high or very high hospital volume based on annual admissions (0–800, 801–1060, 1061–1280,1281+ respectively). Stratification of the continuous variables (age, length of stay in days and number of admissions per year) was based on the median and interquartile range.
Analysis
First, we used chi-square tests to compare demographic, clinical, and hospital factors between patients with and without an AE or ME. Second, we calculated the proportion of hospitalizations with an AE or ME, both overall and separately by each type of event. We then conducted unadjusted and adjusted logistic regression analyses to assess the strength of the relationship between the patient and hospital factors and the AE or ME. Odds ratios were adjusted for all demographic, clinical, and hospital factors and we considered predictors in the adjusted analyses statistically significant at a significance level of <0.05 (2-tailed). Finally, we present box-and-whisker plots to describe the distribution of unadjusted hospital AE and ME rates calculated by using STATA version 14.0. Analyses were performed using Complex Samples analysis in SPSS (IBM Corp, Armonk, NY, USA), version 24 which accommodated the study design’s two-stage proportional sampling and its weighting.
RESULTS
Of the 4,371 discharges reviewed, 48% were of males and 52% were of females. The mean +/−SD age of the discharged patients was 43.5+/−16.9. The mean length of stay was 8.3+/−8.5 days. We found no significant differences between the included and excluded hospitals in the hospital level variables. Discharged patients with an AE were more likely than those without an AE to be older and to have a longer stay, but they were less likely to have suicidality, drug dependence and less likely to be uninsured. We found similar associations for patients who had an ME while hospitalized (Table 1). Overall, AEs occurred during 14.5% of all hospitalizations (Table 2). In 35% of all hospitalizations with an AE, a ME was also identified. The most commonly identified AE were adverse drug events (9.3% of all hospitalizations) and falls (3.9% of all hospitalizations). An ME was identified during 9.0% of all hospitalizations, and medication errors were the most common type (5.7%). In 57% of the hospitalizations with a ME, an AE also occurred.
Table 1.
Characteristics of hospitalizations with and without adverse event or medical error (n=43711).
| Variable | Percentage Adverse Event absent N=3752 (%)1 | Percentage Adverse Event present N=619 (%)1 | p | Percentage Medical Error absent N=3981 (%)1 | Percentage Medical Error present N=390 (%)1 | p |
|---|---|---|---|---|---|---|
| Patient-level factors | ||||||
| Gender | .108 | .077 | ||||
| Female | 52.0 | 56.2 | 52.0 | 58.7 | ||
| Male | 48.0 | 43.8 | 48.0 | 41.3 | ||
| Race | .300 | .187 | ||||
| White | 76.3 | 79.3 | 76.3 | 81.5 | ||
| Non-white | 23.7 | 20.7 | 23.7 | 18.5 | ||
| Age in years | .001 | <.001 | ||||
| 18–30 | 26.6 | 20.3 | 26.3 | 19.4 | ||
| 31–42 | 24.7 | 15.5 | 24.1 | 16.0 | ||
| 43–53 | 25.3 | 20.9 | 24.9 | 22.6 | ||
| 54 or older | 23.4 | 43.3 | 24.8 | 42.0 | ||
| Length of stay in days | <.001 | <.001 | ||||
| 1–3 | 28.0 | 6.3 | 26.2 | 11.4 | ||
| 4–6 | 34.2 | 16.4 | 32.7 | 20.0 | ||
| 7–9 | 17.7 | 20.7 | 18.2 | 17.3 | ||
| 10 or more | 20.1 | 56.6 | 22.8 | 51.3 | ||
| Admission day | .086 | .238 | ||||
| Weekdays | 78.4 | 81.2 | 79.0 | 76.7 | ||
| Weekend | 21.6 | 18.8 | 21.0 | 23.3 | ||
| Principal diagnosis2 | .068 | .122 | ||||
| Mood | 58.3 | 58.2 | 58.6 | 55.3 | ||
| Psychosis | 22.2 | 26.7 | 22.3 | 28.3 | ||
| Other | 19.5 | 15.1 | 19.1 | 16.5 | ||
| Suicidality3 | .001 | .036 | ||||
| Yes | 15.5 | 9.9 | 84.9 | 89.6 | ||
| No | 84.5 | 90.1 | 15.1 | 10.4 | ||
| Drug dependency4 | .001 | .002 | ||||
| Continuous | 8.6 | 4.9 | 8.4 | 4.1 | ||
| Unspecified or episodic | 23.3 | 15.6 | 22.6 | 17.2 | ||
| No or in remission | 68.2 | 79.5 | 68.9 | 78.7 | ||
| Insurance status | .008 | .010 | ||||
| Uninsured | 8.2 | 5.6 | 8.2 | 4.5 | ||
| Commercial | 40.5 | 31.3 | 39.9 | 32.3 | ||
| Medicaid/Medicare | 51.2 | 63.1 | 51.9 | 63.2 | ||
| Hospital-level factors | ||||||
| Teaching status | .064 | .599 | ||||
| Yes | 54.2 | 45.0 | 53.1 | 50.4 | ||
| No | 45.8 | 55.0 | 46.9 | 49.6 | ||
| Urbanity | .729 | .478 | ||||
| Rural | 18.1 | 17.2 | 17.8 | 19.2 | ||
| Urban | 81.9 | 82.8 | 82.2 | 80.8 | ||
| Admissions per year | .058 | .087 | ||||
| Low (0–800) | 29.6 | 32.2 | 29.6 | 33.4 | ||
| Medium (801–1060) | 29.7 | 37.6 | 31.0 | 29.8 | ||
| High (1061–1280) | 22.3 | 14.9 | 21.9 | 13.8 | ||
| Very high (1281 or more) | 18.4 | 15.3 | 17.5 | 22.9 |
All analyses were performed using complex samples analysis, therefore results are presented as percentages.
ICD-9 Psychosis: 295, 297, 298; Mood: 296; Other: all other codes.
ICD-9 Suicidal ideation V62.84 or attempt E950-E959
ICD-9 Drug use: Continuous 303.01, 303.91 304.21–304-.91, 305.01; Unspecified or episodic 303.00, 303.90, 304.20–304.90, 305.20, and 303.02, 303.92, 304.22–304.92, 305.02; In remission: 303.03 303.93, 304.23–304.93, 305.03
Table 2.
Prevalence and rates of patient safety events and medical errors among inpatient discharges (n=43711).
| Event | Hospitalizations with events | Percentage of all hospitalizations | 95% confidence interval |
|---|---|---|---|
| N | % | ||
| Any Adverse Event | 619 | 14.5 | 11.7–17.9 |
| Any Adverse drug event | 398 | 9.3 | 7.1–12.2 |
| Any Patient fall | 160 | 3.9 | 3.2–4.8 |
| Any Other adverse event | 65 | 1.5 | 0.9–2.4 |
| Any Patient assault | 56 | 1.2 | 0.91-.6 |
| Any Patient sexual contact | 42 | 0.9 | 0.7–1.2 |
| Any Patient self-harm or injury | 27 | 0.6 | 0.3–1.1 |
| Any Medical Error | 390 | 9.0 | 7.5–11.0 |
| Any medication error | 249 | 5.7 | 4.3–7.5 |
| Any nondrug errors | 120 | 2.8 | 2.2–3.6 |
| Any contraband1 | 64 | 1.5 | 1.1–2.0 |
| Any elopement1 | 7 | 0.2 | 0.1–0.4 |
Elopement and contraband are errors by proxy.
Table 3 presents the adjusted and unadjusted logistic regression results for patient and hospital factors as predictors of an AE. In the unadjusted model, several variables were significantly associated with a patient–s increased odds of experiencing an AE: (older age (>54); longer stay (four or more days); psychosis as a principal diagnosis (compared with other); absence of suicidality or drug dependence (compared with unspecified or episodic use or continuous use); insurance with Medicare or Medicaid (compared with commercial insurance or uninsured); and low- or medium- volume hospitals (compared with high). After adjustment for all other variables, the variables of age (p=.031), length of stay (p,.001), insurance status (p=.029), and hospital volume (p,.001) were significantly associated with a patient’s odds of experiencing an AE. Specifically, a hospitalization of four or more days was associated with increased odds of an AE (highest odds for 10 or more days; adjusted odds ratio AOR]=11.87), and patients ages 31–42 had lower odds of an AE during hospitalization (AOR=.71) compared with the youngest age group (18–30). Patients seen in low, medium or very high volume hospitals were more likely to have an AE than those seen in high volume hospitals. Compared with patients with Medicaid or Medicare, those with commercial insurance had lower odds of an AE (AOR=.78).
Table 3.
Results of logistic regression analyses for any adverse event during hospitalization (n=4,371).
| Variable | Percentage with adverse event % | 95%confidence interval | UnadjustedOR | 95%confidence interval | Adjusted OR5 | 95%confidence interval |
|---|---|---|---|---|---|---|
| Patient-level factors | ||||||
| Gender | ||||||
| Female | 15.5 | 12.2–19.6 | 1.19 | 0.96–1.47 | 0.99 | 0.81–1.20 |
| Male | 13.4 | 10.8–16.6 | REF | REF | REF | REF |
| Race | ||||||
| White | 14.9 | 11.8–18.7 | 1.19 | 0.84–1.69 | 1.01 | 0.74–1.38 |
| Non-white | 12.8 | 9.6–17.0 | REF | REF | REF | REF |
| Age in years | ||||||
| 18–30 | 11.5 | 9.3–14.1 | REF | REF | REF | REF |
| 31–42 | 9.6 | 8.0–11.6 | 0.82 | 0.66–1.02 | 0.71 | 0.54–0.93 |
| 43–53 | 12.3 | 9.5–15.8 | 1.08 | 0.82–1.43 | 0.82 | 0.60–1.13 |
| 54 or older | 23.9 | 16.6–33.2 | 2.42 | 1.44–4.07 | 1.21 | 0.79–1.88 |
| Length of stay in days | ||||||
| 1–3 | 3.7 | 2.3–5.7 | REF | REF | REF | REF |
| 4–6 | 7.5 | 5.4–10.5 | 2.14 | 1.24–3.72 | 2.12 | 1.11–4.04 |
| 7–9 | 16.6 | 13.2–20.6 | 5.22 | 3.38–8.09 | 5.07 | 3.05–8.42 |
| 10 or more | 32.4 | 25.5–40.1 | 12.58 | 8.38–18.89 | 11.87 | 7.33–19.21 |
| Admission day | ||||||
| Weekdays | 14.7 | 11.5–18.5 | REF | REF | REF | REF |
| Weekend | 12.7 | 9.4–16.9 | 0.84 | 0.69–1.03 | 0.95 | 0.79–1.14 |
| Principal diagnosis2 | ||||||
| Mood | 14.5 | 12.1–17.3 | 1.29 | 0.87–1.90 | 1.24 | 0.91–1.67 |
| Psychosis | 17.0 | 12.5–22.6 | 1.55 | 1.13–2.13 | 0.99 | 0.72–1.37 |
| Other | 11.7 | 7.8–17.2 | REF | REF | REF | REF |
| Suicidality3 | ||||||
| Yes | 9.8 | 8.3–11.5 | 0.60 | 0.46–0.78 | 0.84 | 0.57–1.25 |
| No | 14.5 | 12.1–19.3 | REF | REF | REF | REF |
| Drug dependency4 | ||||||
| Continuous | 8.9 | 6.4–12.2 | 0.49 | 0.36–0.67 | 0.94 | 0.64–1.36 |
| Unspecified or episodic | 10.2 | 8.0–13.0 | 0.58 | 0.39–0.85 | 0.75 | 0.58–0.98 |
| No or in remission | 16.5 | 12.8–21.2 | REF | REF | REF | REF |
| Insurance status | ||||||
| Uninsured | 10.2 | 7.2–14.1 | 0.55 | 0.34–0.89 | 0.79 | 0.59–1.06 |
| Commercial | 11.4 | 10.2–12.7 | 0.63 | 0.44–0.89 | 0.78 | 0.61–0.99 |
| Medicaid/Medicare | 17.0 | 12.0–23.6 | REF | REF | REF | REF |
| Hospital-level factors | ||||||
| Teaching status | ||||||
| Yes | 12.4 | 10.2–14.9 | 0.69 | 0.46–1.03 | 0.96 | 0.64–1.45 |
| No | 17.0 | 12.8–22.2 | REF | REF | REF | REF |
| Urbanity | ||||||
| Rural | 13.9 | 11.3–17.0 | 0.94 | 0.64–1.37 | 1.06 | 0.73–1.54 |
| Urban | 14.7 | 11.3–18.8 | REF | REF | REF | REF |
| Admissions per year | ||||||
| Low (0–800) | 15.6 | 13.2–18.3 | 1.63 | 1.05–2.53 | 1.90 | 1.47–2.45 |
| Medium (801–1060) | 17.7 | 12.0–25.3 | 1.90 | 1.04–3.46 | 2.10 | 1.48–2.99 |
| High (1061–1280) | 10.2 | 7.1–14.4 | REF | REF | REF | REF |
| Very high (1281 or more) | 12.4 | 11.4–13.4 | 1.25 | 0.83–1.87 | 1.36 | 1.01–1.84 |
Adjusted for all patient-level and hospital-level factors.
ICD-9 Psychosis: 295, 297, 298; Mood: 296; Other: all other codes.
ICD-9 Suicidal ideation V62.84 or attempt E950-E959
ICD-9 Drug use: Continuous 303.01, 303.91 304.21–304-.91, 305.01; Unspecified or episodic 303.00, 303.90, 304.20–304.90, 305.20, and 303.02, 303.92, 304.22–304.92, 305.02; In remission: 303.03 303.93, 304.23–304.93, 305.03.
When MEs were included in the adjusted model for AEs, the association was highly significant (AOR=9.07 95%CI (6.14–13.38).
Table 4 presents potential predictors of MEs in the adjusted and unadjusted logistic regression models. In the unadjusted model older age (54 years and older), longer stay (more than six days), insurance with Medicaid or Medicare, (compared with commercial insurance or uninsured) and an absence of drug dependence or suicidality were significantly associated with increased odds of an ME. In the adjusted model, age (p=value=0.006), length of stay (p=value<0.001), admission day of the week (p=value=0.030), urbanity (p=value<0.001) and hospital volume (p=value= 0.005) showed significant associations with a ME. Older age (54 years and older), longer stay (more than six days) and admission during the weekend were associated with an increased odds of a ME. In addition, the odds of an ME during hospitalization were higher for patients seen in rural hospitals (AOR=1.45) or in very-high-volume hospitals (AOR=2.23) (compared with high-volume hospitals).
Table 4.
Results of logistic regression analyses for any medical error during hospitalization (n=4,3711).
| Variable | Percentage with adverse event % | 95% confidence interval | Unadjusted OR | 95% confidence interval | Adjusted OR5 | 95% confidence interval |
|---|---|---|---|---|---|---|
| Patient-level factors | ||||||
| Gender | ||||||
| Female | 10.0 | 8.2–12.2 | 1.32 | 0.97–1.79 | 1.29 | 0.89–1.87 |
| Male | 7.8 | 5.8–10.5 | REF | REF | REF | REF |
| Race | ||||||
| White | 9.5 | 7.6–11.7 | 1.37 | 0.84–2.23 | 1.46 | 0.91–2.34 |
| Non-white | 7.1 | 4.8–10.4 | REF | REF | REF | REF |
| Age in years | ||||||
| 18–30 | 6.8 | 5.5–8.4 | REF | REF | REF | REF |
| 31–42 | 6.2 | 4.4–8.6 | 0.90 | 0.65–1.25 | 0.74 | 0.52–1.07 |
| 43–53 | 8.2 | 6.3–10.6 | 1.23 | 0.89–1.70 | 1.02 | 0.71–1.47 |
| 54 or older | 14.4 | 10.8–18.9 | 2.30 | 1.69–3.12 | 1.32 | 1.04–1.67 |
| Length of stay in days | ||||||
| 1–3 | 4.1 | 2.8–6.0 | REF | REF | REF | REF |
| 4–6 | 5.7 | 4.4–7.4 | 1.41 | 0.96–2.06 | 1.39 | 0.89–2.15 |
| 7–9 | 8.6 | 6.0–12.0 | 2.19 | 1.19–4.03 | 2.41 | 1.35–4.32 |
| 10 or more | 18.2 | 13.5–24.0 | 5.17 | 3.57–7.49 | 4.78 | 3.35–6.81 |
| Admission day | ||||||
| Weekdays | 8.5 | 6.9–10.5 | REF | REF | REF | REF |
| Weekend | 9.7 | 6.9–13.3 | 1.15 | 0.90–1.46 | 1.29 | 1.00–1.67 |
| Principal diagnosis2 | ||||||
| Mood | 8.5 | 7.0–10.4 | 1.09 | 0.70–1.71 | 1.11 | 0.72–1.72 |
| Psychosis | 11.1 | 8.0–15.3 | 1.46 | 0.90–2.36 | 1.21 | 0.70–2.10 |
| Other | 7.9 | 5.2–11.7 | REF | REF | REF | REF |
| Suicidality3 | ||||||
| Yes | 6.4 | 4.8–8.5 | 0.65 | 0.44–0.97 | 1.03 | 0.60–1.77 |
| No | 9.4 | 7.8–11.7 | REF | REF | REF | REF |
| Drug dependency4 | ||||||
| Continuous | 4.6 | 3.0–7.1 | 0.43 | 0.29–0.64 | 0.80 | 0.50–1.28 |
| Unspecified or episodic | 7.0 | 5.1–9.5 | 0.66 | 0.47–0.95 | 0.91 | 0.66–1.25 |
| No or in remission | 10.1 | 8.2–12.5 | REF | REF | REF | REF |
| Insurance status | ||||||
| Uninsured | 5.1 | 2.6–10.0 | 0.46 | 0.23–0.93 | 0.51 | 0.17–1.51 |
| Commercial | 7.3 | 5.8–9.1 | 0.67 | 0.48–0.93 | 0.78 | 0.59–1.03 |
| Medicaid/Medicare | 10.5 | 7.7–14.2 | REF | REF | REF | REF |
| Hospital-level factors | ||||||
| Teaching status | ||||||
| Yes | 8.6 | 6.0–12.1 | 0.90 | 0.58–1.39 | 1.14 | 0.84–1.57 |
| No | 9.5 | 7.8–11.4 | REF | REF | REF | REF |
| Urbanity | ||||||
| Rural | 9.6 | 9.4–9.9 | 1.10 | 0.83–1.45 | 1.45 | 1.19–1.75 |
| Urban | 8.9 | 6.9–11.3 | REF | REF | REF | REF |
| Admissions per year | ||||||
| Low (0–800) | 10.0 | 9.4–10.7 | 1.79 | 0.99–3.24 | 1.60 | 0.96–2.67 |
| Medium (801–1060) | 8.7 | 6.4–11.6 | 1.53 | 0.78–3.00 | 1.51 | 0.93–2.48 |
| High (1061–1280) | 5.9 | 3.3–10.1 | REF | REF | REF | REF |
| Very high (1281 or more) | 11.5 | 7.7–16.7 | 2.08 | 0.99–4.34 | 2.23 | 1.19–4.19 |
Adjusted for all patient-level and hospital-level factors.
ICD-9 Psychosis: 295, 297, 298; Mood: 296; Other: all other codes.
ICD-9 Suicidal ideation V62.84 or attempt E950-E959
ICD-9 Drug use: Continuous 303.01, 303.91 304.21–304-.91, 305.01; Unspecified or episodic 303.00, 303.90, 304.20–304.90, 305.20, and 303.02, 303.92, 304.22–304.92, 305.02; In remission: 303.03 303.93, 304.23–304.93, 305.03.
Figure 1 shows the distribution of facility rates of any AE or ME per 100 admissions for all included acute care general hospitals. The mean hospital rate of AEs per 100 admissions for study hospitals was 14.15 (range 8.12–24.52), and the mean hospital rate of MEs per 100 admissions for study hospitals was 8.94 (range 3.57–15.48).
Figure 1.
Distribution of the rates of events for study hospitals (n=14).
DISCUSSION
By identifying patient- and hospital-level factors that were significantly associated with the odds of an AE or an ME, this study has laid a foundation for further understanding and preventing safety events for patients receiving care in psychiatric units at acute care general hospitals. The findings also help identify hospital system vulnerabilities that can be targeted for patient safety improvements.
Patient-level factors, including longer length of stay, older patient age, admission during the weekend and Medicaid or Medicare insurance compared with commercial insurance status were associated with higher risk of AE or ME.
These findings align with prior research in general hospital units, which has also found that longer length of stay, older patient age, and weekend admission are significantly associated with increased odds of experiencing an AE or ME (20, 21). Length of stay likely increases odds of AE and ME because of a patient’s longer exposure to inpatient care (21). However, it is also possible that a longer stay is a proxy for greater severity of illness, because sicker patients often require more intensive and, consequently, lengthy treatment. Thus, identifying patients who may experience longer stays (e.g., those who have prior involuntary commitments) and implementing additional safety protocols for these high-risk patients, such as more intensive monitoring and continual assessment, could reduce their likelihood of experiencing an AE or ME. Suicidal patients had shorter stays, and after the analysis controlled for amount of time spent in the hospital, suicidality was no longer associated with likelihood of experiencing an AE or ME. Even though patient age patient age is not modifiable, it is useful to know that elderly patients may be at especially high risk of iatrogenic exposure on the psychiatric unit. This finding may prompt exploration of targeted strategies that enhance safety for these high-risk patient populations. Finally, the ‘weekend-effect’ has been identified in other hospital settings to be associated with adverse patient outcomes (22) and is not unique to psychiatric units. One common intervention used in other specialties, and readily adaptable to psychiatry, is to provide 7-day hospital services with experienced staff and access to specialized diagnostics and therapeutics (22). These findings should be used to parallel the processes used in general medical and surgical care to develop interventions reducing for AEs and MEs and for improving the quality of care for psychiatric patients.(5–14)
Our study found that patients seen at hospitals with low, medium, or very high patient volume (compared with high volume) and rural hospitals (compared with urban) were more likely to experience an AE or ME. Operating at overcapacity and with understaffing and higher patient-to-nurse ratios are factors known to be associated with an increase in rates of patient safety events (23). On the other hand, smaller hospitals may be underresourced leading to an increased risk of AE. The optimal equilibrium of staffing and resources to provide safe care may be present in high-volume hospitals. Such equilibrium may be lacking in very-high volume hospitals, where the most ill patients likely present for treatment. Similarly, the findings regarding rural hospitals may point to the need for improvements in the areas of staffing, staff training, and process management. Improving systems of care and structural issues that put patients at risk of experiencing a patient safety event can be instrumental in developing interventions. It is not surprising that organizational factors play a key role in maintaining a safe and therapeutic environment given our understanding of the nature of the hospital milieu and prior research on safety in hospital-based mental health settings (24). One potential intervention, Safewards, is an example of a major innovation that addresses several organizational factors in order to improve the safety of patients on psychiatric wards. Safewards consists of strategies that address six key domains: the staff team, the physical environment, events and relationships outside the hospital grounds, the patient community, patient characteristics, and the regulatory framework (25). In a clustered randomized controlled trial, Safewards was effective in reducing harmful situations (e.g. assaults, self-harm) (26), suggesting that certain aspects of patient safety in inpatient psychiatry can be improved by implementing “systems” or organizational interventions.
Limitations
As with every study, this study had limitations. First, chart review studies may not document the full nature and outcomes of care. However, existing research shows that systematic methods of detecting AEs are 10 times more effective than other detection methods, such as voluntary reporting (27). Second, the use of administrative data to test patient and hospital predictors may have been influenced influenced by the presence of measurement error for clinical characteristics (e.g., diagnosis) that are not collected by validated instruments. Third, we did not have access to data about important covariates, such as comorbid medical issues that may complicate risk for AE and ME and socio-economic status. Fourth, because only one hospital included in this study was a for profit hospital we were unable to examine the potential role of ownership in the occurrence of AE and ME. Fifth, the presence and extent of patient psychological harm were difficult to ascertain in a chart review, a trauma-informed lens that takes these factors into consideration should be employed when designing future research and safety interventions in inpatient psychiatry. Sixth, this study of inpatient psychiatric hospitalizations was conducted at general acute care hospitals and may not be applicable to psychiatric hospitals or longer-term units. For example, patients admitted to general hospitals are likely to have a shorter stay compared with patients admitted to stand-alone psychiatric hospitals. Finally, these findings may not be generalizable to other states given the role of varying state regulations (licensure, policies, insurance mandates, and so forth).
Implications
Essential to patient safety is understanding the underlying patterns of patient and provider factors that are associated with AEs and MEs. Ours is the first large scale study that used an established methodology (17) to examine predictors of AE and ME in inpatient psychiatric units, providing new information about the patient and hospital factors associated with patient safety events among psychiatric inpatients of acute general care hospitals. Extensive inter-hospital variability exists in rates of AE and ME in medical/surgical hospitalizations (20, 28), which suggests that poor-performing hospitals could learn from “high performers” by comparing hospitals systems, identifying gaps and improving care. In general medicine, research has established a framework to lower rates of preventable harm (14) in five steps: measurement, evidence-based care practices, investment in implementation sciences, local ownership and peer learning, aligning and synergizing efforts around common goals and measures. The field of mental health care would do well to adopt a similar framework in which this, and other studies measuring AE and ME, can serve as a first step. From our findings, we can then move to the next step and develop evidence-based practices that address the specific vulnerabilities to patient safety in inpatient psychiatry, by using high-quality interventions focused on improving care paths, such as ‘plan-do-study-act’ cycles (29). These interventions should be targeted to patient groups with the highest risk of experiencing a patient safety event. The effectiveness of improvement interventions could be evaluated by using the recently published roadmap from an international consensus group in the field of psychiatry (30). Future studies should continue the steps along the framework in order to develop comprehensive safety improvements for this vulnerable patient population.
Conclusions
This study examined risk factors for a broad array of safety events in inpatient psychiatric care at acute care general hospitals. The patient and hospital factors that we identified as predicting patient safety events suggest that policies and practices should be targeted at the unit and hospital level to ensure an adequate and safe level of care, during all shifts and at all hospitals regardless of location. By targeting opportunities and strategies to prevent AE and ME in inpatient psychiatry, the field moves one step closer to the end goal of ensuring that psychiatric inpatients receive care in a safe environment.
Supplementary Material
Acknowledgements
This research was supported by grant MH086722 from the National Institute of Mental Health and by the National Patient Safety Foundation.
Footnotes
Disclosures
Dr. X reports receipt of consulting fees from Allergan, Alkermes, Johnson & Johnson, Sage, and Sunovion. The other authors report no financial relationships with commercial interests.
REFERENCES
- 1.Kohn LT, Corrigan J, Donaldson MS: To err is human: building a safer health system. Washington DC: National Academy Press, 1999 [PubMed] [Google Scholar]
- 2.Griffin FA RR: IHI Global Trigger Tool for Measuring Adverse Events (Second Edition). Edited by paper. IISw. Cambridge, MA: Institute for Healthcare Improvement, 2009 [Google Scholar]
- 3.Kizer KW, Stegun MB: Serious Reportable Adverse Events in Health Care; in Advances in Patient Safety: From Research to Implementation (Volume 4: Programs, Tools, and Products). Edited by Henriksen K, Battles JB, Marks ES, et al. Rockville (MD), 2005 [PubMed] [Google Scholar]
- 4.Wu AW, Cavanaugh TA, McPhee SJ, et al. : To tell the truth: ethical and practical issues in disclosing medical mistakes to patients. J Gen Intern Med 12:770–5, 1997 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Spath PL (ed)^(eds): Error Reduction in Healthcare: A Systems Approach to Improving Patient Safety. San Francisco: Jossey-Bass, 2000 [Google Scholar]
- 6.Wachter R, Shojania K: Internal Bleeding: The Truth Behind America’s Terrifying Epidemic of Medical Mistakes. New York City: Rugged Land, 2004 [Google Scholar]
- 7.Institute of Medicine: Patient Safety: Achieving a New Standard for Care. Washington, DC: National Academies Press, 2004 [PubMed] [Google Scholar]
- 8.Millenson ML: Demanding Medical Excellence: Doctors and Accountability in the Information Age. Chicago, IL: University of Chicago Press, 2000 [Google Scholar]
- 9.Bogner MS (ed)^(eds): Human Error in Medicine. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc., 1994 [Google Scholar]
- 10.Rosenthal MM, Sutcliffe KM (ed)^(eds): Medical Error: What Do We Know? What Do We Do? San Francisco: Jossey-Bass, 2002 [Google Scholar]
- 11.Ranji SR, Shojania KG: Implementing patient safety interventions in your hospital: what to try and what to avoid. Med Clin North Am 92:275–93, vii–viii, 2008 [DOI] [PubMed] [Google Scholar]
- 12.Hosford SB: Hospital progress in reducing error: the impact of external interventions. Hosp Top 86:9–19, 2008 [DOI] [PubMed] [Google Scholar]
- 13.Brennan TA, Gawande A, Thomas E, et al. : Accidental deaths, saved lives, and improved quality. N Engl J Med 353:1405–9, 2005 [DOI] [PubMed] [Google Scholar]
- 14.Pronovost PJ, Cleeman JI, Wright D, et al. : Fifteen years after To Err is Human: a success story to learn from. BMJ Qual Saf 25:396–9, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Joint Commission on the Accreditation of Healthcare Organizations: Preventing adverse events in behavioral health care. Oakbrook Terrace, IL, 1999 [Google Scholar]
- 16.Poster EC, Pelletier LR, Kay K: A retrospective cohort study of falls in a psychiatric inpatient setting. Hosp Community Psychiatry 42:714–20, 1991 [DOI] [PubMed] [Google Scholar]
- 17.Marcus SC, Hermann RC, Frankel MR, et al. : Safety of Psychiatric Inpatients at the Veterans Health Administration. Psychiatric services:appi. ps. 201700224, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mark TL, Stranges E, Levit KR: Using Healthcare Cost and Utilization Project State Inpatient Database and Medicare Cost Reports Data to Determine the Number of Psychiatric Discharges from Psychiatric Units of Community Hospitals Agency for Healthcare Research and Quality (AHRQ), 2010 [Google Scholar]
- 19.Hiatt HH, Barnes BA, Brennan TA, et al. : A study of medical injury and medical malpractice. The New England journal of medicine 321:480–4, 1989 [DOI] [PubMed] [Google Scholar]
- 20.Naylor R: Risk factors predisposing to adverse events and medication errors; in Medication errors: Lessons for education and healthcare. Oxford, UK: Radcliffe Medical Press, 2002 [Google Scholar]
- 21.Bates DW, Miller EB, Cullen DJ, et al. : Patient risk factors for adverse drug events in hospitalized patients. ADE Prevention Study Group. Archives of internal medicine 159:2553–60, 1999 [DOI] [PubMed] [Google Scholar]
- 22.Walker AS, Mason A, Quan TP, et al. : Mortality risks associated with emergency admissions during weekends and public holidays: an analysis of electronic health records. Lancet, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Weissman JS, Rothschild JM, Bendavid E, et al. : Hospital workload and adverse events. Med Care 45:448–55, 2007 [DOI] [PubMed] [Google Scholar]
- 24.True G, Frasso R, Cullen SW, et al. : Adverse events in veterans affairs inpatient psychiatric units: Staff perspectives on contributing and protective factors. General hospital psychiatry 48:65–71, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bowers L, Alexander J, Bilgin H, et al. : Safewards: the empirical basis of the model and a critical appraisal. J Psychiatr Ment Health Nurs 21:354–64, 2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bowers L, James K, Quirk A, et al. : Reducing conflict and containment rates on acute psychiatric wards: The Safewards cluster randomised controlled trial. International journal of nursing studies 52:1412–22, 2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Classen DC, Resar R, Griffin F, et al. : ‘Global trigger tool’ shows that adverse events in hospitals may be ten times greater than previously measured. Health affairs (Project Hope) 30:581–9, 2011 [DOI] [PubMed] [Google Scholar]
- 28.Leape LL: The preventability of medical injury; in Human Error in Medicine. Edited by Bogner MS. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc., 1994 [Google Scholar]
- 29.Etchells E, Ho M, Shojania KG: Value of small sample sizes in rapid-cycle quality improvement projects. BMJ Qual Saf, 2015 [DOI] [PubMed] [Google Scholar]
- 30.Kilbourne AM, Beck K, Spaeth-Rublee B, et al. : Measuring and improving the quality of mental health care: a global perspective. World psychiatry : official journal of the World Psychiatric Association 17:30–8, 2018 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.