Essays & Theses

A human factors approach to understanding patient safety during pediatric cardiac surgery

Description
A human factors approach to understanding patient safety during pediatric cardiac surgery
Published
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A human factors approach to understanding patient safety during pediatric cardiac surgery Cynthia Galvan a  , Emile A. Bacha a  , Julie Mohr   b , Paul Barach c, * a   Pediatric and Congenital Cardiac Surgery, Section of Cardiothoracic Surgery, University of Chicago, IL, USA  b  Department of Medicine, University of Chicago, IL, USA c  Department of Anesthesiology, Jackson Memorial Hospital, and Center for Patient Safety, University of Miami, North Wing 109 1611 NW 12th Ave, Miami, FL 33136, USA Available online 17 March 2005 Abstract Pediatric cardiac surgery–a highly complex, low-error-tolerant field with patients who are normally sick and require complex correctionsat a very early age–has long recognized the need to study human factors and their relationship to medical outcomes. The current study uses amultidisciplinary pediatric cardiac team at a high-volume institution ( N 300 cases per year) to study patient safety practices using humanfactors methods including a retrospective chart review, safety culture analysis, and observations of the team in the operating room.Preliminary results from observations include examples of major and minor events (ranked by potential patient consequences) compensatedor uncompensated by clinical responses to such events, active and/or latent communication break-downs, and internal and externaldistractions. The culture of care was felt by the providers to be adequate and safety was not felt to be a problem despite the adverse eventsnoted. Major events occurred most frequently during surgical correction and/or bypass run (  N  =16) and post-bypass period (  N  =17). Minor events occurred most frequently during the anesthesia period (  N  =46) and during surgical correction and/or bypass run (  N  =63). Internal andexternal distractions occurred most frequently during the surgical correction and/or bypass run (  N  =167). The majority of clinical responses to problematic events observed occurred as a result of cognitive recognition by practitioners. There were no intra-operative deaths observed. No-harm events, near misses, and adverse events were observed and examples are provided. Future directions include linking observations tomedical outcomes, video-recording operations to analyze technical adverse events or near misses, and surveying team members of major andminor events and clinical responses to these events. D  2005 Elsevier Ireland Ltd. All rights reserved.  Keywords:  Pediatric cardiac surgery; Safety culture; Human factors; Adverse events; Near misses; No harm events; Adverse outcomes 1. Introduction 1.1. Medical errors, adverse events, and system failures In 2000, the Institute of Medicine reported 44,000– 98,000 people die every year in US hospitals as a result of medical errors [1]. Along with national recognition and media attention, the report   To Err is Human  focused thehealth care community on the need to address patient safety.The report suggested that   b the decentralized and fragmentednature of the health care delivery system  Q  , or what theytermed a  b non-system  Q  , contributes to unsafe conditions for  patients and impedes safety improvement efforts [1]. Some researchers believe deaths due to medical errors can be ashigh as the third leading cause of death [2].The Harvard Medical Practice (HMP) [3] Study foundthat, in 4% of admissions in New York State, patientswere unintentionally harmed and 14% died partly as aresult of their treatment. Similarly, an Australian studyfound a 16.6% adverse event rate, with a similar number of serious events compared to the HMP study [4]. Both studies also reported that nearly half the events wereassociated with a surgical intervention, such as technique-related complications, wound infections, or post-operative bleeding [5]. 1058-9813/$ - see front matter   D  2005 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ppedcard.2004.12.001* Corresponding author. Tel.: +1 305 585 6102; fax: +1 305 585 8359.  E-mail address:  pbarach@med.miami.edu (P. Barach).Progress in Pediatric Cardiology 20 (2005) 13–20www.elsevier.com/locate/ppedcard  Most often, it is not a single error that leads to an adverseevent. Rather, it is a succession of events that, while minor in isolation, lead to an adverse event if they occur together.Reason’s  b Swiss Cheese Model  Q   depicts an organizationalaccident and its contributing factors [6]. This modelsuggests there are  b holes  Q   in the system that are dormant elements that lead to errors or failure (or non-existence) of systems to prevent these errors. When they eventually lineup, they allow the error trajectory to harm the patient.Based on a recent study [7] interviewing surgeons fromthree teaching hospitals showing that 33% of recalledadverse events resulted in permanent disability, 77% wererelated to an invasive or surgical procedure, and 66%happened during the intra-operative phase of a procedure,we chose to focus this study on the intra-operative period. Ina majority of the incidents reported in the study, systemfactors including inexperience or lack of competence in asurgical task, communication breakdowns among personnel,fatigue or excessive work load, and emergency surgical careled to system failures [7]. 1.2. Pediatric cardiac surgery as a highly complex, low-error-tolerant medical field  Cardiac surgeons have long recognized the need to studyhuman factors and its relationship to outcomes [8,9]. Pediatric cardiac surgery in particular is an ideal field for human factors study because it is both complex and has alow-error-tolerance [10,11]. It encompasses many complex  procedures that are highly dependent upon a sophisticatedorganizational structure, coordinated efforts of multipleindividuals working as a team, and high levels of cognitiveand technical performance. High-risk populations, such asneonates and infants in particular, exhibit a fragile physi-ology, which can be very unforgiving [12– 14]. Recent  advances have allowed for marked improvement in out-comes and also for more successful early complete repairs,often performed during the neonatal period in low-birthneonates [15 – 17]. This has increased the complexity and low-error-tolerance even further. However, while the pub-lished mortality rates for children older than 1 year undergoing open heart surgery is in the range of 3.5%[12,13], mortality rates for neonatal repairs are still in therange of 10–40%, varying by the underlying lesion, and thetype of repair  [12,14]. Several factors have been linked to the poor outcomes in pediatric cardiac surgery, includinginstitution and surgeon-specific volumes [12,18 – 20], com-  plexity of cases [21], and systems failures. The preventable  pediatric cardiac deaths detailed in the Bristol RoyalInfirmary Inquiry and the Manitoba Inquiry reports bothrecognized the importance of human factors and systemsresearch in improving the pediatric cardiac surgical out-comes [22 – 24]. The Manitoba Pediatric Cardiac Surgery Inquest found that   b serious organizational and personnel problems experienced by the Health Sciences Center’sPediatric Cardiac Surgery Program during 1993 andt hroughout 1994 contributed to the deaths of these children  Q  [24].A methodology was developed to study the human, team,and organizational factors that influence the outcomes of  pediatric cardiac surgery, based on a theory of organiza-tional accident causation [6]. Using methodology described below, the present study attempts to identify and analyzehuman factors associated with the practice of pediatriccardiac surgery. We chose a multidisciplinary approach toidentify and analyze active failures, latent factors, andresponses to problematic events. We report preliminaryresults of a larger study that links the impact of humanfactors and clinical outcomes. 2. Methods 2.1. The methodology for this study included three elementswhich are summarized below2.1.1. Retrospective chart review Chart reviews, when used in combination with other investigative methods, can contribute to the understanding of latent and act ive errors [25]. Using a previously supported methodology[4,5,26 – 28],wescreened500chartsina2stage chart review protocol and found 60 adverse events since theinception of the program in 2000. The retrospective chart review revealed multiple adverse events, latent errors, andactive errors during care. These events included unplannedextubations, carotid artery punctures, infections, etc. Theseevents were collected, analyzed, and used to inform our  prospective observations in the operating room. 2.1.2. Safety culture assessment survey We designed a questionnaire based on our review of theliterature and our own observations [29,40] (Fig. 1). All members of the Pediatric Cardiac Surgery (PCS-surgery,anesthesia, perfusionists, cardiology, intensive care, nurs-ing) team were asked for their opinions about the safetyculture in their hospital. The goal was to survey the culturein the beginning stages of this project, then do so again at the completion of the project to measure any differencesfound at the completion of this project. The introduction of an observer into the operating room can be thought of as an b intervention  Q   —for example, we have noticed an increasedawareness of safety issues since beginning the observations. 2.1.3. Observations in the operating room Real-time observations of the PCS team in the in theoperating room (OR) are currently underway. Observations begin at the inception of anesthesia and end with the patient hand-off to the cardiac intensive care unit. The observationsincluded how the PCS teams respond and recover fromevents. PCS teams are observed and team member interactions are recorded by a trained observer. A list of complex high-risk surgeries to be observed was chosen C. Galvan et al. / Progress in Pediatric Cardiology 20 (2005) 13–20 14  (Table 1). We follow the surgical flow and record all major and minor events as well as clinical responses to suchevents. Major and minor events ar e defined based on potential consequences to the patient  [9]. 2.2. The pediatric cardiac surgery (PCS) program and  patient and procedural data The PCS program at the University of Chicago started inearly 2000 and currently performs an average of 300 openand closed cases per year. In addition to local patients, 10%are transferred from out-of-state locations for complexsurgeries. Neonates and infants comprise 50% of cases,while adults with congenital heart disease comprise 7% of cases. All cases are extensively discussed during a weeklymultidisciplinary conference. The PCS team consists of two board-certified pediatric cardiac surgeons, as well asdedicated anesthesiologists, cardiologists, perfusionists,operating room nurses, and pediatric cardiac intensivists.Rapid-deployment extra-corporeal membrane oxygenation(ECMO) is used when needed. The 30-day and hospitalmortality rates were 2.0% and 3.6% for 2002 and 2.5% and4.1% for 2003, respectively. Awell-documented prospectivedatabase is kept. Patient variables collected for correlationwith observations of human factors in the OR are derivedfrom those collected in the PCS database [30]. The pediatric cardiac nomenclature and outcomes analysis used in thisstudy are widely accepted methods, validated by the Societyof Thoracic Surgeons [21,30]. The complexity scores are derived from the Aristotle  R  scoring system [21]. Analysis of data is underway and currently unavailable. 2.3. Definitions Traditionally, emerging research fields have difficulties innomenclature consensus. Table 2 shows a list of definitionswe chose to use, based on extensive review of the literature.Based on the definitions, we categorized events (medicalinterventions) with potential harm as no harm events, near misses, adverse events, and adverse outcomes (Fig. 1). 2.4. Informed consent  Full institutional review board approval was attained.Written consent was discussed and acquired from PCS teammembers, patients, and parents or guardians. HIPAAauthorization was gained from patients and parents or guardians. 3. Results This paper focuses on the outcomes from the observa-tions. Results from the chart review and safety survey will Table 1List of procedures that are prospectively observed !  Any neonatal ( b 29 days of age) procedure (open and closed) !  Arterial switch operation !  Stage I Norwood !  Aortopulmonary shunt  !  Cavopulmonary shunt  !  Fontan procedure !  Repair any type of tetralogy of fallot  !  Repair complete atrioventricular canal !  Ross procedureTable 2DefinitionsDefinitionsActive failures/errors— errors and procedural violations which arecommitted by the medical team and have an immediate impact on patient safety [9].Adverse events—undesirable and unintended incidents in care that mayresult in adverse outcomes or may require additional care efforts tothwart adverse outcome. Criteria of disability, injur y, and causalityattributed to medical care and not disease process [27].Adverse outcomes—undesirable and unintended outcomes of care such asdeath, disability, or temporary disability [34].Clinical responses to er rors —measures used to counter an unintendedconsequence in care [34].Dangerous situations—where both human and latent factors exist that create a hazard increasing the risk of harm [34].Disability—a temporary or permanent impairment of physical (includingdisfigurement) or mental function or increased length of stay and/or economic loss (even in the absence of such impairment) [28].Latent factors/errors—system weaknesses that result from decisions made by the higher management in the organization, by regulators,governments, designers, and manufacturers. Latent conditions lead toweaknesses in the organization’s defenses, thus increasing the likelihoodthat when active failures occur they will combine with existing preconditions, breach the system’s defenses, and result in an accident.Examples include system defects such as poor design, incorrect installation, faulty maintenance, poor purchasing decisions, andinadequate staffing [25,37]. Major event—is one that is thought to have serious consequences for thesafety of the patient if not corrected (i.e., laceration of an artery) [9]. Minor event—is one that disrupted the  b surgical flow  Q   of the procedure but which, in isolation, were not expected to have serious consequencesfor the safety of  the patient if not corrected (i.e., handing error by thescrub nurse) [9]. Medical error—the failure of a planned action to be completed as intendedor the use of a wrong action to achieve an aim. Errors can include problems in practice, products, procedures, and systems [38].  Near miss—event in which the unwanted consequence is prevented because there was a recovery by identification and correction of thefailure. Such a recovery could be by a planned or unplanned barrier  [39]. No harm events—events that have occurred but resulted in no actual harmalthough the potential for harm may been present. Lack of harm may bedue to the robust nature of human physiology or pure luck. (An exampleof such a benign event would be the issuing of an ABO incompatibleunit of blood for a patient, but the unit was not transfused and wasreturned to the blood bank.) [36]Preventable adverse events—a subset of adverse outcomes that are judgedt o have been avoidable if appropriate and reasonable steps had been taken[36].Safety culture—the product of the individual and group values, attitudes, perceptions, competencies, and patterns of behavior that determine t hecommitment to an organization’s health and safety management  [6]. Sentinel events— events in which death or serious harm to a patient hasoccurred [36]. C. Galvan et al. / Progress in Pediatric Cardiology 20 (2005) 13–20  15   be reported separately. As noted above, the chart reviewrevealed errors, near misses, and adverse events in the peri-operative treatment of these children. Fig. 1 shows thecategorization of medical interventions with potential harmas: no harm events, near misses, adverse events, and adverseoutcomes. Preliminary results from the safety surveysuggest a lack of awareness to patient safety hazard as wellas lack of awareness of the various ways to keep childrenfrom being harmed in the operating room. 3.1. OR observations During a 3-month period, September 1st to November 30th, 2003, 80 cardiac surgery cases were performed. Thehospital mortality rate was 2.5% (  N  =2). Twenty-six cases fit the study criteria. Twenty-two cases were observed andincluded in the results. Two cases were not observed due toobserver unavailability, and two cases had incompleteobservational data and were not included. Table 3 lists thediagnoses, procedures, and patient complexity scores. Thecauses of death included multiple organ failure and sepsisfor the two cases of mortality at post-op days 20 and 49,respectively. 3.2. Major and minor events All adverse events, near misses, and no harm events werecoded using 1 of 6 domains: pre-operative, anesthesia, pre- bypass, during surgical correction, post-bypass, and trans- port to intensive care unit. Events were described ascompensated or uncompensated. An event was consideredcompensated when an appropriate clinical response wasused to attenuate the consequences from the event (Table 4). There were no uncompensated major events in anydomain and no intra-operative adverse outcomes or deathswere observed. However, no harm events, near misses, andrecovery from adverse events were observed. All major events were compensated with the majority occurring duringsurgical correction and/or bypass run (  N  =16) and post- bypass period (  N  =17). 3.2.1. Major events Compensated major events most represented in alldomains were associated with bleeding, heart rhythmabnormalities, ventilation issues, and myocardial protectionduring bypass (cardioplegia). Clinical recovery from harmfulevents included mostly cognitive recognition by practi- Medical InterventionPotential Harm EventPure Luck orPhysiologyBarriers and/orSecondary MechanismsTemporary HarmCould the event been avoidedif appropriate and reasonablesteps had been taken?Preventableadverse eventUnpreventableadverse eventNoYes Analysis of Harm Event  Permanent Harm No HarmNear MissAdverse Outcome Harm (Adverse Event) Fig. 1. Categorization of medical interventions with potential harm as: no harm events, near misses, adverse events, and adverse outcomes. C. Galvan et al. / Progress in Pediatric Cardiology 20 (2005) 13–20 16  tioners.Inonecase,transfusionbestpracticeguidelinesaidedthe practitioners in recognizing a potentially harmful event. 3.2.2. Minor events The majority of minor events were recorded during theanesthesia period (  N  =46) and during surgical correction(  N  =63). Internal (i.e., alarm, phone, pager, and conversa-tions) and external (i.e., noise and people other than teammembers walking in and out of OR) distractions wererecorded as minor events but were listed separately. Themajority of internal and external distractions occurredduring surgical correction (  N  =167). Active and latent communication breakdowns were recorded as minor events,unless they had severe consequences for the team and the patients. Below are three examples.Case 1: No harm event During patient transport from the operating room to theCardiac Intensive Care Unit (CICU), the oxygen bottleattached to the bed fell when the transport team turned acorner at a swift pace. The oxygen bottle began to leak rapidly and its tight connection was lost. A new bottle wasobtained and the transport team continued with no patient sequelae.Case 2: Near miss with clinical response to correct situationDuring transesophageal echo (TEE) probe insertion, whichoccurred with some difficulty, the surgeon noticed the patient’s chest was no longer rising and no longer beingventilated. The anesthesiologist checked the endotrachealtube and it appeared that during TEE probe insertion the patient was extubated. The patient was re-intubated. Thesurgeon had the team take the chest roll out from under the patientandthenreattempttheTEEprobeinsertion.Theprobewentinsmoothly.ThesurgeonadvisedtheORteamfromthat moment on, the back roll was to be placed under the patient  after   the TEE probe insertion. Table 3List of diagnosis, procedures, and complexity scores of included subjectsCase # Diagnosis Age/weight (kg)Procedure(s)Outcome Aristotle (R)complexity score Clinical profile of observed patients 1 D-TGA NB/3.1 kg ASO well 10.02 AS, subAS, endocardial fibroelastosis,aortic CoA, hypoplastic arch, shock  NB/2.6 kg –Ross-Konno-aortic arch augmentation died POD 49 20.5 –ECMO3 Aortic CoA, hypoplastic arch, Turner’s NB/2.8 kg Aortic arch augmentation on CPB well 12.04 D-TGA (Taussig-Bing), aortic CoA NB/2.8 kg ASO, VSD closure, aortic arch augment well 18.05 HLHS, s/p stage I Norwood 6 months/5 kg BDG well 9.56 TOF/CAVC NB/3.1 kg Mod. BT shunt well 10.37 Truncus arteriosus/left PAPVR NB/2.5 kg –Rastelli procedure, repair PAPVR 12.0 –ECMO8 Same patient as above –RV-PA conduit revision died POD 20 12.59 Tricuspid atresia NB/2.6 kg Mod. BT shunt well 13.310 HLHS, s/p stage I Norwood 2 years/11 kg Fontan well 14.511 AS/AI, tunnel subaortic stenosis 3 years/15 kg Aortic valvuloplasty, resectionsubaortic tunnelwell 10.012 Pulmonary atresia/intact ventricular septum 1 year/8 kg BDG well 8.813 DORV/CAVC 2 years/13 kg Fontan well 14.514 TOF/absent left pulmonary artery NB/3 kg Central shunt to left pulmonary artery well 6.815 DORV/subAS/pulmonary HTN 7 months/5 kg –DKS/BDG 14.016 Same patient as above –BDG take-down, BT shunt well 14.317 TOF/CAVC NB/3 kg Mod. BT shunt well 11.318 Tricuspid atresia 2 years/15 kg Fontan well 11.019 AS/AI 14 years/60 kg Ross procedure well 12.320 Aortic CoA, hypoplastic Arch NB/2.8 kg Extended end-end repair well 16.021 TAPVR to coronary sinus NB/3 kg repair well 9.022 Interrupted aortic arch/VSD NB/3 kg repair well 15.8 Clinical profile of patients fitting the study criteria but not included  23 HLHS NB/3 kg Hybrid stage I well 1324 AS/AI 15 years/60 kg Ross procedure well 10.325 TOF 6 months/6 kg Complete repair well 8.026 TOF NB/3.3 kg Mod. BT shunt well 13.3AI: aortic insufficiency; AS: aortic stenosis; ASO: arterial switch operation; BDG: bi-directional Glenn (shunt); BT: Blalock-Taussig (shunt); Co-arc:coarctation of the aorta; CAVC: complete atrioventricular canal; D-TGA: dextra (rotated)-transposition of the great vessels; CPB: cardiopulmonary bypass;DORF: double outlet right ventricle; DKS: Damus Kaye Stansel; ECMO: extra-corporeal membrane oxygenation; EFE: endocardial fibroelastosis; HTN:hypertension; HLHS: hypoplastic left heart syndrome; IAA: interrupted aortic arch; Mod. BT: modified Blalock-Taussig (shunt); NB: newborn; PA: pulmonaryartery; PAPVR: partial anomalous pulmonary venous return; RV: right ventricle; sp: status post; SubAS: sub-aortic stenosis; TAPVR: total anomalous pulmonary venous return; TOF: tetralogy of fallot; VSD: ventricular septal defect. C. Galvan et al. / Progress in Pediatric Cardiology 20 (2005) 13–20  17
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks