You are here

Article Text

Article menu

Download PDFPDF

Original research
The morbidity and mortality conference as an adverse event surveillance tool in a paediatric intensive care unit
  1. Christina L Cifra1,
  2. Kareen L Jones2,
  3. Judith Ascenzi3,
  4. Utpal S Bhalala2,
  5. Melania M Bembea2,
  6. James C Fackler2,
  7. Marlene R Miller4
  1. 1Department of Pediatrics, University of Iowa, Iowa City, Iowa, USA
  2. 2Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University, Baltimore, Maryland, USA
  3. 3Pediatric Intensive Care Unit, Johns Hopkins Hospital, Baltimore, Maryland, USA
  4. 4Department of Pediatrics, Johns Hopkins University, Baltimore, Maryland, USA
  1. Correspondence to Dr Christina L Cifra, Department of Pediatrics, University of Iowa, 200 Hawkins Drive 2JCP, Iowa City, IA 52242, USA; christina-cifra@uiowa.edu

Abstract

Objective To determine if standardised chart review applied to records of patients discussed at a paediatric intensive care unit (PICU) morbidity and mortality conference (MMC) yields additional or different information regarding safety event occurrence and characteristics.

Design Retrospective record review.

Setting Single tertiary referral PICU in Baltimore, Maryland, USA.

Participants 96 patients discussed at the PICU MMC over 14 months (November 2011–December 2012).

Main outcome measures Safety events and their characteristics (medical error category, severity and preventability).

Results A total of 275 safety events were identified through the MMC and/or chart review. The MMC identified 131 (48%) events, 53 (19%) of which were identified through the MMC alone. After chart review was performed, an additional 144 (52%) events were identified. 78 (28%) events were identified through both. High severity adverse events potentially contributing to permanent harm or death were more likely to be identified through both the MMC and chart review (47%) compared with either alone. The MMC alone identified more near-misses (21%) and preventable events (96%) compared with chart review alone or both MMC and chart review. Although chart review alone helped to identify many healthcare-associated infections, medication errors and sedation/pain control issues not elicited through the MMC, the MMC alone identified more communication errors and workflow problems. The MMC alone also identified 40% of all diagnostic errors, which would not have been discovered otherwise despite chart review by itself identifying 50% of such misdiagnoses.

Conclusions Standardised chart review applied to records of patients discussed at a PICU MMC identified significantly more safety events not initially discovered through the MMC. However, the MMC was superior to chart review in identifying broader problems such as communication errors, workflow issues and certain diagnostic errors not captured by chart review, which can potentially affect many aspects of care.

  • Adverse events, epidemiology and detection
  • Critical care
  • Medical error, measurement/epidemiology
  • Morbidity and mortality rounds
  • Paediatrics

https://doi.org/10.1136/bmjqs-2014-003000

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Children are particularly vulnerable to healthcare-associated errors due to multiple factors, with the most serious errors occurring more often in the complex environment of the paediatric intensive care unit (PICU).1 ,2 Retrospective chart reviews in the PICU report that 59%–62% of PICU patients suffered at least one adverse event, with 36%–45% of these deemed preventable.3 ,4

    Given this, PICUs need to effectively survey for patient safety events to be able to identify and prioritise problems and measure progress with interventions. Our recent survey of 84 PICUs (both academic and private) across the USA showed that 100% used voluntary incident reporting as a primary means of safety event detection, augmented mostly by targeted hospital infection control surveillance (98%), while less than 50% used periodic manual or electronic chart reviews or other methods of patient safety monitoring.5 The common utilisation of incident reporting reflects the general practice of hospitals across the country.6

    Prior work has shown that incident reporting alone does not provide a complete picture of a unit7 or a hospital's safety events.8–12 In fact, disproportionate dependence on only one method of surveillance may divert focus from other important unrecognised safety concerns.10 Indeed, many authors have advocated for the adoption of multiple methods of detecting patient safety events with the goal of characterising the full range of issues that need to be addressed. However, adoption of multi-method surveillance has been sparse.13 ,14

    In recent years, many investigators have demonstrated that a structured, systems-oriented morbidity and mortality conference (MMC) can improve reporting, identification and analysis of adverse events.15–17 Because of this, plus the observation that 94% of PICUs we surveyed already conduct regular MMCs,5 we put forth that the MMC might serve as an effective and accessible adjunct adverse event surveillance method. Previous work in adverse event epidemiology, particularly in the PICU,3 ,4 have used standardised chart review as a primary method of event detection; thus, we sought to compare how the MMC performs with regard to safety event identification and characterisation in relation to chart review.

    We hypothesised that the MMC is an effective complementary method of adverse event surveillance in the PICU due to its ability to detect severe but preventable safety events that involve broader elements of care as compared with chart review. The objective of this study was to determine if standardised chart review applied to records of patients discussed at a PICU MMC yields additional or different information regarding safety event occurrence and characteristics.

    Methods

    Study environment and context

    The Johns Hopkins Children's Center PICU is a 36-bed, Level 1 unit18 accommodating ∼1700 admissions per year, of patients 0–21 years of age, and serves as a tertiary referral centre for many local and international centres. It is staffed by one to three paediatric critical care physicians and three fellows for 24 h each day divided into two 12-h shifts. Patients include a wide variety of both surgical and medical cases, including patients undergoing solid organ transplants, requiring extracorporeal membrane oxygenation, or needing intensive trauma and burn care. Cardiac ICU patients are admitted in the general PICU but cohorted with trained cardiac PICU nurses.

    Our PICU conducts patient safety event surveillance using many approaches, including a hospital-wide, voluntary, electronic incident reporting network. Johns Hopkins Hospital uses University HealthSystem Consortium's Patient Safety Net,19 where any staff member can access the system at any time through an electronic portal in the institution-wide intranet. The system receives an average of 300–350 incident reports per year from our PICU. We also obtain infection surveillance data through the Department of Hospital Epidemiology and Infection Control. Aside from this, we independently conduct active surveillance for catheter-associated urinary tract infections and for central line-associated bloodstream infections (CLABSI) as part of the Children's Hospital Association PICU CLABSI Collaborative.20 We do not routinely conduct periodic chart reviews for safety event surveillance.

    Our PICU holds twice-monthly MMCs where three to five cases are discussed each session. Cases can be suggested for discussion by any PICU clinician, including non-physicians. Almost all cases suggested are discussed at conference. There are no set criteria for selecting cases, though we do discuss all mortalities. Our experience shows that clinicians usually suggest cases where patient harm or the potential for patient harm occurred due to possibly remediable actions. There is a separate conference where we discuss cases of interest due to unusual presentations, clinical course, pathophysiology or management. Prior to the MMC, cases for discussion are collated into an MMC meeting agenda outlining the patients’ demographic information with brief descriptions of the hospital course, reason for discussion including any adverse events reported and autopsy results if applicable, as reported by the clinician suggesting the case for discussion. No standard preparation (including reviews of patient charts) is done prior to the MMC. Participants, including clinicians who were primarily involved in the patients’ care, largely discuss the case from memory.

    Data sources and record review

    All patients discussed at the PICU MMC during the study period were included. We excluded patients who were discussed but not actually admitted to the PICU. MMC agendas, patient charts and incident report records were used to identify and characterise safety events; however, we focused mainly on safety events identified through the MMC and/or standardised chart review. This study was approved by the Johns Hopkins Institutional Review Board in Baltimore, Maryland, USA.

    We conducted a systematic retrospective record review of all patients discussed at the PICU MMC using a modified ‘trigger tool’ review methodology previously used by Agarwal et al3 for identifying and classifying safety events. We integrated a review of each patient's case summary (this includes MMC-reported safety events) as documented in the MMC agenda within the process of reviewing the same patient's medical records. We first identified all safety events discussed at the MMC and attempted to determine if the chart contains documentation of the same event. We also recorded all other safety events we identified from the chart review that were not discussed at the MMC. Incident report records for the same patients during the same time period were reviewed separately.

    All reviewers were trained in the review methodology. First-tier review was performed by two physician reviewers (CLC and KLJ), both senior paediatric critical care fellows, who reviewed all three data sources independently. Afterwards, any discrepancies in judgment were discussed, with an attempt to reach consensus between the two reviewers. If consensus was not achieved, items were elevated to a second-tier review by a paediatric critical care attending physician (USB) and senior PICU nurse (JA), who then determined final judgments.

    We defined the following terms as such: (1) safety event (ie, medical error): an act of commission (doing something wrong) or omission (failing to do the right thing) related to healthcare management that leads to or has the potential to lead to an adverse event, and encompasses both adverse events and near-misses;21 (2) near-miss: an act of commission or omission related to healthcare management that by chance did not lead to an adverse event but has the potential to lead to an adverse event;22 and (3) adverse event: an unintended physical or mental injury that resulted in prolongation of hospital stay, temporary or permanent disability, or death, and was caused by healthcare management rather than the patient's disease.3 ,23

    For each identified safety event, we classified them as either a near-miss or adverse event per the definitions above. We further classified each event by medical error category (modified for the PICU setting), severity and preventability using previously published error classification systems.24 ,25 Error categories were adapted from the United States Agency for Healthcare Research and Quality's (AHRQ) publicly available Common Format guidelines,24 which were then modified for applicability to the PICU setting. Severity was graded according to the modified National Coordinating Council for Medication Error Reporting and Prevention Index for Categorizing Errors (box 1).3 ,25 Preventability was graded according to the AHRQ Common Format guidelines (box 2).24

    Box 1 Grading of adverse event severity*

    • Category E: contributed to or resulted in temporary harm to the patient and required intervention

    • Category F: contributed to or resulted in temporary harm to the patient and required initial or prolonged hospitalisation

    • Category G: contributed to or resulted in permanent patient harm

    • Category H: required intervention to sustain life

    • Category I: contributed to or resulted in the patient's death

    *Adapted from the National Coordinating Council for Medication Error Reporting and Prevention Index for Categorizing Errors, 200125

    Box 2 Grading of adverse event preventability*

    • Category 1: almost certainly could have been prevented

    • Category 2: likely could have been prevented

    • Category 3: likely could not have been prevented

    • Category 4: almost certainly could not have been prevented

    *Adapted from the Agency for Healthcare Research and Quality Common Format24

    Statistical analysis

    Descriptive data analysis was performed with medians, IQRs and proportions reported as appropriate. We compared characteristics of all PICU admissions with that of the study cohort for the same time period using K-sample equality of medians test or Fisher's exact test. Severity and preventability of safety events identified through the MMC alone, chart review alone or through both were compared using χ2 goodness-of-fit or Fisher's exact test. A p value of 0.05 was considered statistically significant. Statistical analysis was performed using STATA V.12.1 (StataCorp, College Station, Texas, USA, 2011).

    Results

    Study cohort characteristics

    There were 98 patients discussed at the PICU MMC (5.4% of all PICU admissions) during the 14-month study period from November 2011 to December 2012. We excluded two patients both of whom died in the emergency department and were never admitted to the PICU. Overall, 89 of these remaining 96 patients (93%) had at least one identified safety event.

    Characteristics of the patients included in the study compared with all PICU admissions for the same time period are outlined in table 1. There were no statistically significant differences in demographic and clinical characteristics between the study cohort patients and all PICU admissions except for a longer length of stay (9.5 vs 2 days), and a lesser proportion of (mostly elective) surgical cases (35% vs 54%) among patients discussed at the MMC. All deaths during the time period were discussed, which comprised 2.5% of all admissions and 46% of all study patients.

    View this table:
    Table 1

    Demographic and clinical characteristics of patients discussed at the PICU MMC compared with all admissions

    Safety event characteristics

    Safety event identification by surveillance tool

    Among patients discussed at the PICU MMC, a total of 377 safety events were identified through the MMC, chart review and incident reporting. Chart review identified most events (59%), followed by the MMC (35%) and incident reporting (31%). There was very little overlap between the three methods of surveillance, with only 2% of safety events identified by all three methods. There was also minimal overlap between incident reporting and the MMC (1%). There was no overlap at all between incident reporting and chart review, with each identifying a completely different set of safety events. The MMC however, overlapped considerably with chart review (18%) (figure 1).

    Figure 1
    Figure 1

    Safety event identification by surveillance tool. This diagram illustrates the relative proportions of safety events identified by each surveillance tool among patients discussed at the PICU MMC. The area of each rectangle represents the surveillance tool's relative contribution to safety event detection. Overlapping areas between rectangles represent safety events detected by two or more surveillance tools as indicated. ‘Total’ refers to the total number of events identified by each tool, alone or in combination with other tools, whereas ‘only’ refers to the proportion of events solely identified by the specific tool. PICU, paediatric intensive care unit; MMC, morbidity and mortality conference; IR, incident reporting.

    The MMC versus chart review

    In table 2, we summarised the characteristics of safety events identified through the PICU MMC and/or standardised chart review, comparing each tool's ability to detect and characterise safety events in the same patient cohort.

    View this table:
    Table 2

    Summary of safety event characteristics identified through the paediatric intensive care unit MMC and/or chart review*

    Overall, a total of 275 safety events were identified through the MMC and/or chart review, most of which were adverse events (94%). The MMC identified 131 (48%) events, 53 (19%) of which were identified through the MMC alone. However, when chart review was applied to the same study cohort's records, an additional 144 (52%) events were identified. A total of 78 (28%) events were identified through both the MMC and chart review.

    The majority of adverse events identified (179, 65%) were classified in the less severe categories E or F. These were events judged to have caused temporary harm that required intervention or initial/prolonged hospitalisation. Chart review alone identified slightly more low severity events (75%) compared with the MMC alone (59%) or both MMC and chart review (51%). In all, 80 (29%) adverse events were classified in the more severe categories G, H or I. These were events judged to have caused permanent harm, required life-sustaining interventions or caused death. These high severity adverse events were more likely to be identified through both the MMC and chart review (47%) compared with either alone. The MMC alone more commonly identified near-misses (21%), events that had the potential to cause patient harm but did not.

    The majority of safety events identified (72%) were classified as preventable. The MMC alone identified a higher proportion of preventable events (96%) compared with chart review alone (69%) or both MMC and chart review (63%).

    In table 3, we summarised the different medical error categories identified through the PICU MMC and/or standardised chart review, comparing each tool's ability to characterise different types of safety events in the same patient cohort. Table 2 also shows the top three most common error categories identified by each. One safety event may be classified into more than one category; thus, there were more categories counted than there were safety events identified (345 vs 275).

    View this table:
    Table 3

    Medical error categories as identified through the paediatric intensive care unit MMC and/or chart review*,†

    Overall, medication errors (18%), healthcare-associated infections (16%) and communication/teamwork errors (15%) were most commonly identified. The application of chart review to the records of patients discussed at the MMC helped to identify majority of healthcare-associated infections (71%), medication errors (56%) and sedation/pain control issues (85%) not elicited through the MMC. However, the MMC alone identified almost all communication errors (83%) and workflow problems (65%) not readily identified through chart review. Both chart review and the MMC separately identified diagnostic errors (50% and 40%, respectively).

    Discussion

    Our study showed that among all safety events identified by the MMC and/or standardised chart review, the MMC identified approximately half of all events (48%), with around a fifth (19%) identified by the MMC alone. Applying chart review to the records of patients discussed at the MMC yielded additional safety events that were not discussed at conference, comprising the other half (52%) of total events. This is not a surprising finding, as our work also found that chart review yielded the largest proportion of safety events for our cohort (59%). This ability to detect a significant volume of adverse events has made chart review popular among safety researchers, becoming widely used as the surveillance tool of choice in many scholarly works on adverse event epidemiology,2 ,3 ,26 including the original landmark report, ‘To Err is Human’ by the Institute of Medicine.27 This emphasises the importance of conducting at least periodic chart reviews to surface these errors; however, many PICUs do not do so because of its resource-intensive nature.28 Our study also showed that compared with the MMC and both MMC and chart review, chart review alone identified mainly low severity events (75%). High severity events causing significant patient harm or death were more likely to be identified by both the MMC and chart review (47%) compared with either alone, as such especially harmful incidents are usually identified by more than one surveillance method.

    It is well established in the literature that different methods of adverse event detection yield different types of adverse events depending on the surveillance tool used.8 ,10 ,14 We demonstrated this in our study, showing that among patients discussed at the PICU MMC, only 2% of safety events were identified by all three methods used: the MMC, incident reporting and chart review. Because of this, many authors have advocated for the adoption of multiple methods of detecting patient safety events with the goal of characterising the full range of issues that need to be addressed.7 ,8 ,10

    Our prior work has shown that most PICUs use incident reporting to surface adverse events, and some conduct additional periodic manual or electronic chart reviews.5 Most PICUs also hold regular MMCs;5 however, it was unknown if the adverse events discussed during these MMCs add meaningfully to make a more comprehensive picture of a unit's adverse events. In this study, we demonstrated that the MMC is indeed a necessary and complementary tool for this purpose.

    The MMC identified more near-misses (21%) and preventable events (96%) than either chart review alone or both the MMC and chart review combined. This implies that clinicians are suggesting for discussion cases with actionable errors, as is the expected nature of volunteered cases since presumably it is more desirable to discuss errors with potential solutions. Thus, the MMC acts as a venue for clinicians to parse out and prioritise errors for which discussion yields potential beneficial action. Near-misses also tend to not be recorded in the chart since no patient harm occurred, and instead are reported in the MMC to be discussed as an example of a possibly substantial error averted. Thus, if we relied solely on chart review to identify safety events, these opportunities to improve may be missed.29

    A striking difference between the characteristics of safety events identified by the MMC versus chart review was the markedly different types of errors that each detected. Performing chart review on records of patients discussed at the MMC did not necessarily corroborate the safety events around which discussion at the MMC was focused. Instead, most of the time, chart review yielded a whole different set of safety events not related to the events discussed at the MMC. However, there were certain types of events, such as diagnostic errors, that were confirmed by chart review if the reviewer had prior knowledge of its occurrence, but would have been difficult to ascertain from the chart without direction.

    Chart review tended to identify discrete, easily identifiable and quantifiable events such as medication errors, healthcare-associated infections and sedation (usually oversedation) or pain control issues. In this manner, chart documentation and incident reporting are similar, wherein reported events are those that usually can be documented with ease (unambiguous incident with a clear what, where and when), or are mandated to be documented by institutional regulatory policies.30 Some of these types of safety events identified by chart review are also identified by other means. For our PICU, medication errors were usually detected through incident reporting, while active infection surveillance detected certain types of healthcare-associated infections such as CLABSI and catheter-associated urinary tract infections.

    In comparison, the MMC identified broader, more sweeping concerns not as easily documented in the chart. Topping the list were communication/teamwork errors and workflow issues, most of which were identified solely through the MMC. Furthermore, we also noted that the MMC alone identified 40% of all diagnostic errors, which would not have been discovered otherwise even though chart review alone identified slightly more of the same (50%). We surmise that this may be because chart review uncovers diagnostic errors much later after the fact (eg, autopsy reports inconsistent with clinical diagnoses), while the MMC identifies diagnostic errors discovered earlier in living patients by clinicians. This finding illustrates the utility of the MMC in augmenting identification of misdiagnoses beyond autopsy results, perhaps leading to important insights for prevention.

    These types of safety events identified by the MMC were also usually the main reasons for suggesting a case for discussion, as the problems were more complicated and multi-faceted, and needed consideration from multiple perspectives if effective interventions were to be implemented. Although any clinician can report cases for discussion at our MMC, our experience shows that most suggested cases usually come from physicians. This illustrates a possible reason for the differential in types of safety events identified by chart review and the MMC. Nurses, concerned with minute-to-minute care, usually document most medication and procedural errors in the chart or through incident reports, whereas physicians, engaged more with the long term trajectory of care, usually report wider concerns focused on diagnosis and management at the MMC.7 Another possible reason for the discrepancy may be the staff's distinction of the different avenues for error reporting; thus, more discrete, less severe events with clear cause-and-effect are documented in the chart or filed as an incident report, whereas larger events with far-ranging repercussions requiring multidisciplinary staff discussion are reported through the MMC.

    These findings exhibit the importance of the MMC in eliciting general safety events that can potentially affect many aspects of care and, if addressed, can effect more widespread positive change. An added benefit of the MMC is that, in contrast to other adverse event surveillance tools, it provides an opportunity for indepth, systems-oriented, multidisciplinary analysis of events that may lead to effective and well-accepted solutions.15 ,31

    Our study has several limitations. This is a single centre retrospective study. Retrospective record reviews may be subject to recall and outcome bias. We were only able to include patients discussed at the PICU MMC, with no comparison review of patients not discussed. Generalisability of findings may be limited to settings similar to the study context. Additionally, for voluntary forms of adverse event surveillance, it has been show that the amount and quality of safety events captured vary with the unit's safety and reporting culture.32–34 This will need to be kept in mind when extrapolating study findings to individual PICUs. It is also important to note if a particular PICU integrates incident reporting and/or chart review with their MMC (eg, events uncovered by chart review are discussed at the MMC, or safety events discussed at the MMC are subsequently recorded in patient charts), in which case our findings would be difficult to apply.

    Conclusions

    Standardised chart review applied to records of patients discussed at a PICU MMC helped to identify significantly more safety events not initially identified through discussion at the MMC. However, the MMC is superior to chart review in identifying broad and possibly pervasive problems that can potentially affect many aspects of care, such as communication errors and workflow problems. Thus, the MMC is a necessary complementary adverse event surveillance tool that can augment patient safety monitoring by surfacing more general, expansive problems in care delivery that if addressed, can effect more widespread positive change. Future research is needed in methods of consolidating, analysing and prioritising safety event data obtained from multiple sources to create an integrated model of adverse event surveillance and response, and the effects of such a model on improving patient safety.

    Acknowledgments

    The authors are grateful for the contributions of Eileen Kasda, MHS, from the Johns Hopkins Patient Safety Office and Jeffrey Natterman, Esq, from the Johns Hopkins Medicine Legal Department who assisted in obtaining and interpreting incident report data. We also thank Carol Thompson, MS, from the Johns Hopkins School of Public Health Biostatistics Center and Erik Westlund, MSc, from the Johns Hopkins Department of Sociology who assisted in statistical analysis.

    References

    1. ↵
      Steering Committee on Quality Improvement and Management and Committee on Hospital Care. Policy statement--principles of pediatric patient safety: reducing harm due to medical care. Pediatrics 2011;127:1199–210.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Matlow AG,
      2. Baker GR,
      3. Flintoft V,
      4. et al
      . Adverse events among children in Canadian hospitals: the Canadian Paediatric Adverse Events Study. CMAJ 2012;184:E709–18.
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. Agarwal S,
      2. Classen D,
      3. Larsen G,
      4. et al
      . Prevalence of adverse events in pediatric intensive care units in the United States. Pediatr Crit Care Med 2010;11:568–78.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Larsen GY,
      2. Donaldson AE,
      3. Parker HB,
      4. et al
      . Preventable harm occurring to critically ill children. Pediatr Crit Care Med 2007;8:331–6.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Cifra CL,
      2. Bembea MM,
      3. Fackler JC,
      4. et al
      . The morbidity and mortality conference in pediatric intensive care units in the United States: a national survey. Crit Care Med. In press.
    6. ↵
      1. Farley DO,
      2. Haviland A,
      3. Champagne S,
      4. et al
      . Adverse-event-reporting practices by US hospitals: results of a national survey. Qual Saf Health Care 2008;17:416–23.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Silas R,
      2. Tibballs J
      . Adverse events and comparison of systematic and voluntary reporting from a paediatric intensive care unit. Qual Saf Health Care 2010;19:568–71.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Thomas EJ,
      2. Petersen LA
      . Measuring errors and adverse events in health care. J Gen Intern Med 2003;18:61–7.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Olsen S,
      2. Neale G,
      3. Schwab K,
      4. et al
      . Hospital staff should use more than one method to detect adverse events and potential adverse events: incident reporting, pharmacist surveillance and local real-time record review may all have a place. Qual Saf Health Care 2007;16:40–4.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Levtzion-Korach O,
      2. Frankel A,
      3. Alcalai H,
      4. et al
      . Integrating incident data from five reporting systems to assess patient safety: making sense of the elephant. Jt Comm J Qual Patient Saf 2010;36:402–10.
      OpenUrlPubMed
    11. ↵
      1. Michel P
      . Strengths and weaknesses of available methods for assessing the nature and scale of harm caused by the health system: literature review. Geneva: World Health Organization (WHO), 2003: 59. http://www.who.int/patientsafety/research/P_Michel_Report_Final_version.pdf (accessed 16 Jun 2014).
    12. ↵
      1. Sun F
      . Monitoring patient safety problems. In: Making healthcare safer II: an updated critical analysis of the evidence for patient safety practices. Rockville, MD: Agency for Healthcare Research and Quality, 2013. http://www.ncbi.nlm.nih.gov/books/NBK133411/ (accessed 8 Nov 2013).
    13. ↵
      1. Shojania KG
      . The frustrating case of incident-reporting systems. Qual Saf Health Care 2008;17:400–2.
      OpenUrlFREE Full Text
    14. ↵
      1. Shojania KG
      . The elephant of patient safety: what you see depends on how you look. Jt Comm J Qual Patient Saf 2010;36:399–401.
      OpenUrlPubMed
    15. ↵
      1. Deis JN,
      2. Smith KM,
      3. Warren MD,
      4. et al
      . Transforming the morbidity and mortality conference into an instrument for systemwide improvement. In: Henriksen K, Battles JB, Keyes MA, et al., eds. Advances in patient safety: new directions and alternative approaches (Vol. 2: Culture and Redesign). Rockville, MD: Agency for Healthcare Research and Quality, 2008. http://www.ncbi.nlm.nih.gov/books/NBK43710/ (accessed 8 Nov 2013).
    16. ↵
      1. McDonnell C,
      2. Laxer RM,
      3. Roy WL
      . Redesigning a morbidity and mortality program in a university-affiliated pediatric anesthesia department. Jt Comm J Qual Patient Saf 2010;36:117–25.
      OpenUrlPubMed
    17. ↵
      1. McVeigh TP,
      2. Waters PS,
      3. Murphy R,
      4. et al
      . Increasing reporting of adverse events to improve the educational value of the morbidity and mortality conference. J Am Coll Surg 2013;216:50–6.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Rosenberg DI,
      2. Moss MM
      . Guidelines and levels of care for pediatric intensive care units. Pediatrics 2004;114:1114–25.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      UHC. https://www.uhc.edu/UHCPSN.htm (accessed 21 Jan 2014).
    20. ↵
      1. Miller MR,
      2. Niedner MF,
      3. Huskins WC,
      4. et al
      . Reducing PICU central line-associated bloodstream infections: 3-year results. Pediatrics 2011;128:e1077–83.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      AHRQ Patient Safety Network. http://www.psnet.ahrq.gov/popup_glossary.aspx?name=error (accessed 20 Jan 2014).
    22. ↵
      AHRQ Patient Safety Network. http://www.psnet.ahrq.gov/popup_glossary.aspx?name=nearmiss (accessed 20 Jan 2014).
    23. ↵
      1. Wachter R
      . Understanding patient safety. 2nd edn. McGraw-Hill Professional, 2012.
    24. ↵
      AHRQ. AHRQ Common Formats. http://www.pso.ahrq.gov/formats/commonfmt.htm (accessed 20 Jan 2014).
    25. ↵
      NCC MERP—Medication Error Category Index. http://www.nccmerp.org/medErrorCatIndex.html (accessed 6 Dec 2012).
    26. ↵
      1. Landrigan CP,
      2. Parry GJ,
      3. Bones CB,
      4. et al
      . Temporal trends in rates of patient harm resulting from medical care. N Engl J Med 2010;363:2124–34.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Kohn LT,
      2. Corrigan JM,
      3. Donaldson MS
      , eds. To err is human: building a safer health system. Committee on Quality of Health Care in America, Institute of Medicine. Washington DC: The National Academies Press, 2000.
    28. ↵
      1. Stockwell DC,
      2. Kane-Gill SL
      . Developing a patient safety surveillance system to identify adverse events in the intensive care unit. Crit Care Med 2010;38:S117–25.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Marks CM,
      2. Kasda E,
      3. Paine L,
      4. et al
      . ‘That was a close call’: endorsing a broad definition of near misses in health care. Jt Comm J Qual Patient Saf 2013;39:475–9.
      OpenUrlPubMed
    30. ↵
      1. Levinson DR
      . Hospital incident reporting systems do not capture most patient harm [OEI-06-09-00091]. Washington DC: Department of Health and Human Services, Office of Inspector General, 2012: 42. http://oig.hhs.gov/oei/reports/oei-06-09-00091.asp (accessed 16 Jun 2014).
    31. ↵
      1. Deshpande JK,
      2. Throop PG,
      3. Slayton JM
      . Standardization of case reviews (morbidity and mortality rounds) promotes patient safety. Pediatr Clin North Am 2012;59:1307–15.
      OpenUrlCrossRefPubMed
    32. ↵
      1. Hutchinson A,
      2. Young TA,
      3. Cooper KL,
      4. et al
      . Trends in healthcare incident reporting and relationship to safety and quality data in acute hospitals: results from the National Reporting and Learning System. Qual Saf Health Care 2009;18:5–10.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Stockwell DC,
      2. Slonim AD
      . Quality and safety in the intensive care unit. J Intensive Care Med 2006;21:199–210.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Abstoss KM,
      2. Shaw BE,
      3. Owens TA,
      4. et al
      . Increasing medication error reporting rates while reducing harm through simultaneous cultural and system-level interventions in an intensive care unit. BMJ Qual Saf 2011;20:914–22.
      OpenUrlAbstract/FREE Full Text
    View Abstract

    Footnotes

    • Contributors CLC conceptualised and designed the study, performed first-tier retrospective record reviews, coordinated all activities pertinent to the work, drafted the initial manuscript, performed statistical analysis, and approved the final manuscript as submitted. KLJ performed first-tier retrospective record reviews, participated in the initial analyses, reviewed and revised the manuscript, and approved the final manuscript as submitted. JA participated as a second-tier reviewer, participated in the initial analyses, reviewed and revised the manuscript, and approved the final manuscript as submitted. USB contributed to the conception and design of the study, participated as a second-tier reviewer, participated in the initial analyses, reviewed and revised the manuscript, and approved the final manuscript as submitted. MMB, JCF and MRM contributed to the conception and design of the study, participated in the initial analyses, reviewed and revised the manuscript, and approved the final manuscript as submitted. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

    • Funding Ms Thompson's assistance was supported by the National Center for Research Resources and the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health through Grant Number 1UL1TR001079.

    • Competing interests None.

    • Ethics approval Johns Hopkins Institutional Review Board.

    • Provenance and peer review Not commissioned; externally peer reviewed.