Reducing preventable birth injuries and liability claims through evidence-based care, enhanced teamwork

Transcription

1 Reducing preventable birth injuries and liability claims through evidence-based care, enhanced teamwork Premier Perinatal Safety Initiative Results Summary Released May 2014

2 FINDINGS Focused, consistent adherence to delivering defined clinical care and usingg disciplinedd teamwork and communications processes resulted in a reduction in the incidencee of perinatall harm, and a corresponding reduction in liability claims and costs. SUMMARY OF RESULTS The results of this studyy demonstrated that organized quality improvement efforts coordinated across diverse hospitals in multiple healthcare systems can reduce perinatal harms and related liability costs resulting from adverse events. 14 hospitals participating in the Premier Perinatal Safety Initiative (PPSI) reduced their total adverse events (AOI harm scale) impacting mothers and neonates by 14.4% over this five year quality improvement project, outperforming a set of comparison hospitals. Academic hospitals averaged improvement of 8.8%, while non-academic hospitals averaged improvement of 18.3%.. Participating hospitals lowered PSI 17 Birth Trauma incidents by 15.7% overall, with academic hospitals improving by 30.2% and non-academic hospitals averaging improvement of 3.8%. Incidents of birth Hypoxia and Asphyxia were reduced against baseline by 37.6%, withh academic hospitals averaging improvement of 43.8% and non-academics at 29.7%. Neonatal birth brain injuries arising from these types of incidents were found too average $ million in loss payouts in overall medical malpractice claims closed, indicating that preventionn of just four of these adverse events could save approximately $5 million in liability costs and more in long term care costss and patient suffering. Hospitals also reduced the frequency of new malpractice claims filed on deliveries during PPSI and paid less to close claims compared to baseline. Participating hospitals improved scoring on a High Reliability Safety Assessment overall by 26% compared to baseline, with all hospitals meeting Industry Standards Compliance by project end. The overall final score of 89% was one point short of the highest blue ribbon level of High Reliability Performer. Focus on team communication and simulation may have also improved other perinatal services which were not a focus of the PPSI, such as improving VBAC by 12.8% over the final three years and lowering primary cesarean section rate by two percent-reversing a trend of increasing annually at 5.5% before PPSI. 2 Premier Perinatal Safety Initiative

3 Background The vast majority of U.S. births result in healthy infants and mothers. But when a rare serious adversee event occurs it is devastating for infants, mothers, families and care providers. Ongoing care of infants injured at birth also places significant financial exposure on parents, care providers, insurance companies and public agencies. Though some of these rare events are unavoidable, triggered by pre-existing conditions or other risk factors, estimates suggest 30 percent may be preventable 1. One study of liability claims found that the majority of preventable obstetrical adversee events weree associated with communication gaps between healthcare providers 2, an indicator that effective team- can be building and communication protocols effectivee in preventing perinatal harm. The harm occurring to mothers or babies can lead to court cases with enormous liability verdicts and settlements. Multiple cases of newborns suffering brain injuries have led to verdicts of more than $20 million in the past two years. One 2013 study by a hospital insurer estimated that OB injuries are involved in 43 percent of the total malpractice cases exceeding $5 million in losss payout. 3 Consequently, over half of the typical hospital ss risk management budget is spent in the labor and delivery area. 4 In a recent survey, 58 percent of obstetrician-gynecologists said they changed how they practiced due to the risk or fear of being sued. 5 Maternal admissions with complications are about twice as costly as stays without complications, and admissions with pregnancy and delivery-related complications account for $17.4 billion in annual U.S. hospital costs. 6 In addition, hospital stayss with pregnancy-related complications tend to be longer ( days) than without complicating conditions (1.9 days) ). 7 Givenn the high volume of deliveries, and the potentially devastating emotional and financial impact of birth-related adverse events, optimizing Obstetrical care is a must. Liability risk, costs impacting patient access to care The ncreasing risk of medical liability, coupled with the high cost of liability insurance, has Ob- to how and even whether they offer birthing services. Approximately 9 percent of all U.S. birthss suffer an adverse event. 1 1 Gyns and hospitalss nationwide making changes One recent study estimated the rate of malpractice claims filed for perinatal harms at 5.64 per 10,000 births, 12 but the rate varies basedd on the sample of hospitals examined. Another study estimated the claim rate in 2011 was nearly double for academicc hospitals than community hospitals in another insurance pool (six vs. three cases per 10,000) ). 13 According to a September survey by the American Congress of Obstetricians and Gynecologists (ACOG) 14 of more than 9,0000 board-certified members of ACOG, Among practicing Obstetricians, 13.5 percent stopped performing vaginal birth after cesarean delivery (VBAC), eight percent decreased the number of total deliveries, and five percent discontinuedd the practice of Obstetrics altogether. Among practicing gynecologists, 12 percent decreased gynecologic surgical procedures with five percent no longer performing major gynecologic surgery. Previous ACOG surveys suggested that, on average, 90 percent of ACOG board-certifiedd members have been sued while practicing, and Ob-Gyns can expect to be sued 2.7 times on average during their career. 15 With approximately 4 million babies born annually in the U.S., it s no surprise that childbirth is the number one reason for hospital admissions. 8,9 Nearly a quarter of all hospital discharges are either mothers or newborns, 10 second only to cardiac care in terms of total cost in the U.S. Medicaid provides funding for nearly 40 percent of all births. Premier Perinatal Safety Initiative 3

4 The University of Alabama reported maternity service hospitals declinedd from 58 in 1980 to 32 in January 2011, with similar reports from New York City, where at least five hospitals have ceased providing maternity care since Studies completed since 2012 have demonstrated perinatal lawsuits consume the most time from judges, 19 take longer to resolve than other claims, and cost significantly more to settle and defend than all other cases. 20 Annual reports from two states show the significant financial costs related to perinatal malpractice claims. In Ohio 21 a seven year analysis ( ) found the average loss paid by hospitals for OB department claims was $881,104 (three times higher than all other losses), with defense costs averaging $107,672 (higher than all other departments). Physicians with Obstetrics specialty averaged $431,126 per loss paid (1.5 times higher than all others) and defense costs averaged $59,910 per claim. Typically, physicians have less insurance than hospitals, so their loss averages tend to be lower. In the event a physician ss policy allows defense costs to erode the limits available to pay loss, or the limits are only to $1 million or less, their policy may be insufficient to satisfy high demands in perinatal cases, causing the plaintiff to focus on the hospital to pay more. In Washington state 22 an analysis of closed claims similarly showed labor and delivery claims against hospitals averaged $895,536 and defense expensess averaged $ 93,402. Physicians specializing in Obstetrics averaged a loss of $507,199 (highest of all specialties) and defense costs of $ 59,559 per case. About the Premier Perinatal Safety Initiativee ( ) Premier, Inc. launched the Premierr Perinatal Safety Initiative (PPSI)) in 2008 as one of the largest and most sophisticated perinatal improvement initiatives of its kind. When the initiative was launched, little empirical evidence existed about the effectiveness of methods to improve perinatal patient safety, and few scientifically-designed initiatives existed studying the relationship between improved patient safety and reduced Obstetrical events. While other healthcaree providers and organizations have since implemented Obstetrical improvement projects,, the ability to measure or gauge success has proven to be difficult. Some of those projects were off short duration, unable to effectivelyy track long term change in delivery of care or subsequent malpractice claims which result years later. Other projects were designed around a single hospital or healthcare system with employedd physicians, so the dynamics of organizing independent providers were not fully appreciated. The PPSI offered a longer horizon to evaluate overall impact of interventions, as well as examinedd the interplay among diverse and unaffiliated hospitals and their independent providers. The goalss of the PPSI were: 1.) Lower the incidence of certain infrequent, though serious, injuries that could result in a wide range of harmful outcomes, including birth asphyxia and permanent neurologic disability; 2.) Better define preventable perinatal harm; 3.) Identify care practices that can result in improved outcomes; and 4.) Measure the financial value of these care improvements, including evaluation of whether harm reductions lead to fewer liability claims and less costly pay-outs. 4 Premier Perinatal Safety Initiative

5 The PPSI project management team included Premierr Insurance Management Services, Inc. (PIMS); Fairview Health Services; KTConnolly & Associates, LLC; the University Of Minnesota School Of Public Health; the School for the Sciencee of Health Care Delivery; Arizona State University; and the National Perinatal Information Center (NPIC). Participants included 14 of the country's leading hospitals (See Figure 1 next page), large and small, teaching and non-teaching, system-baseand non-employed physicians. Twelve states and stand-alone, with employed were represented, and approximately 349,000 babies were deliveredd at the project hospitals over the collaborative s five years. A group of comparator hospitals was developed for PPSI to measure potential influences on outcomes from environmental factors, as well as to benchmark with the results from the participating hospitals. The comparator hospitalss were selected from NPIC membership, in orderr to ensure similar comparative harm data from 2006 to 2012 was available, and to best match certain characteristics with the participating hospitals, such as location and birth volume. The participating hospitals were insured by American Excess Insurance Exchange, RRG (AEIX), which paid for the first phase of the PPSI. AEIX provided malpractice claims data to PIMS. Similar hospitals in AEIX, which were not in PPSI, were monitored and their claims and financial data were used in the aggregate to compare claim results to the PPSI participating hospitals. Premier Perinatal Safety Initiative 5

6 Perinatal Safety Initiative Participants Figure 1 Illinois UnityPoint Health Methodist (Peoria) * S Kentucky Baptist Health Louisville, part of Baptist Health (Louisville) M Massachusetts Baystate Medical Center, part of Baystate Health (Springfield) * M Minnesota Fairview Ridges Hospital, part of Fairview Health Services (Burnsville) M University of Minnesota Medical Center, Fairview, part of Fairview Health Services (Minneapolis) * S New Mexico Presbyterian Hospital, part of Presbyterian Healthcare Services (Albuquerque) M Ohioo Bethesda North Hospital, part of TriHealth (Cincinnati) * M Good Samaritan Hospital, part of TriHealth (Cincinnati) * L Summa Akron City Hospital, Summa Health System (Akron) * M Tennessee Indian Path Medical Center, part of Mountain States Health Alliance (Kingsport) S Texas Texas Health Harris Methodist Hospital Fort Worth (Fort Worth) M Texas Health Presbyterian Hospital Dallas (Dallas) L Washington PeaceHealth St. Joseph Medical Center (Bellingham) S Wisconsin Aurora West Allis Medical Center, Aurora Health Care (West Allis) M (S) Small birth volume births; (M) Medium birth volume births; (L) Greater than or equal to 5000 births; (*) Academic teaching status 6 Premier Perinatal Safety Initiative

7 Figure 2 Scope of the PPSI The PPSI was unique due to the broad mix of participating hospitals and their birth volumes, and in the availability of and access to detailed information on Obstetrical liability claims. For these reasons, the lessons learned from the PPSI are valuable and applicable to a wide range of hospitals across the country. The focus of the PPSI was to help participating hospitals prevent five significant clinical issues. Research has shown that these recurring issues are responsible for the majority of perinatal harm and associated costs, including Obstetric professional liability claims. 23 These include: Failure to recognize an infant in distress; Failure to initiate a timely cesarean birth; Failure to properly resuscitate a depressed baby; Inappropriate use of labor-inducing drugs; and Inappropriate use of vacuum or forceps. The initiative also evaluated participant claim/ /lawsuit information to determine if lowering the rate of harm lowers the number of claims filed or the amounts paid to resolvee them. Data concerning the participating hospitals' malpractice claims and payouts for perinatal injury have been and will continue to be compiled and analyzed by PIMS, and compared against similar hospitals from liability insurer, AEIX, which is managed by PIMS. PPSII Timeline The PPSI was comprised of three phases (Figure 2). The Baseline Phase consisted of the retrospective collection of harm (outcome) data and malpractice claims data from births in 2006 and 2007, to establish a baseline of performance. During Phase I, which ran from January 2008 through December 2010, hospital teams met and then implemented clinical interventions while actively working on performance and perinatal safety improvements involving approximately 145,000 births. Phase II, which added interventions and collected data from January 2011 through December 2012, was funded by a grant from the Department of Health & Human Services' Agency for Healthcare Research and Quality (AHRQ). The grant was awarded through PPSI participant Fairview Health Services. Interventions from Phase I continued in Phase 2, with the addition of in-situ simulation training using TeamSTEPPS principles providedd at each hospital site (for a list of interventions, see Appendix page 22). Approximately 200,000 births at 14 hospitalss occurred in Phase II, along with 164,800 deliveries studied at the group of comparator hospitals. Data collection ended December 31, 2012, with analysis throughout 2013 by Fairview Health Services, the University of Minnesota School of Public Health, PIMS, the School for the Science of Health Care Delivery, Arizona State University and the National Perinatal Information Center. Premier Perinatal Safety Initiative 7

8 Figure 3 Two powerful tools to improve perinatal care Experts have found that miscommunicationn and other preventable factors contribute to the harm sustained by some mothers and newborns during labor and delivery. 24 They believe that many of these injuries can be prevented through development of high reliability healthcare teams. Leveraging knowledge gained from previous initiatives, including a Premier/Institute for Healthcaree Improvement (IHI)/Ascension Health collaboration, participating hospitals used two powerful methods to create high reliability healthcare teams: 1.) Increased adherence to evidence-basedd care bundles; and 2.) Enhanced communication and teamwork. Increased adherence to evidence-basedd care bundles Research suggests that grouping essential processes together in care bundles helps clinical staff remember to take all the necessary steps to provide optimal care to every patient, every time. 25 Although many hospitals have long followed some or all of these individual care practices to improve outcomes in childbirth, the key is consistently using all the practices in concert. These care bundles (Figure 3) were developed by IHI in adherence to published best practices and national standards established by leading maternity healthcare groups, such as the ACOG and Association of Women s Health, Obstetric and Neonatal Nurses (AWHONN). However, long term outcome data supporting these bundles has not been previously collected andd published. Requirements for each bundle differ, but have the same objective: standardize clinical processes and reduce variation in practices. Care bundle achievement is scored in an all-or- none fashion; the care team must provide all elements of care in the bundle to be given credit for its use. Compliance was measured via monthly chart audits. Forr example, the goal of one care bundle was to reduce the risks associated with augmenting labor, particularly the use of the drug oxytocin to initiate labor or to accelerate a slow labor. This bundle has four elements which must be used consistently. If a team neglected to document an estimate of the fetal weight before administering the medication, it would not receive credit forr the bundle, even if team members successfully implemented the three other elements. Thee use of oxytocin is commonly cited in medical malpractice claims as causing distress to the neonate, and the support for this in recent literature 26 makes it importantt for providers to establish oxytocin safety protocols and enforce them even in the absence of a bundle. 8 Premier Perinatal Safety Initiative

9 Bundle resultss The goal of achieving 100 percent bundle compliance was achieved by multiple hospitals over several quarterly periods. The overall average compliance rate improved from 38% to 81% between 2008 and 2012, an increase of 81 percent. Significant improvement began from the outset, and gradually plateaued in the 2 nd quarter of Surveys demonstrated areas the teams needed to overcome to push results higher, including competing tasks prevented fulll engagement and certain aspects of bundles were difficult to achieve uniformly, such as getting the estimated fetal weight. Successful teams overcame turnover of key leadership and staff as a challenge to achieving continued compliance over five years. Teams had to develop strategies to get bundle elements fulfilled by outside physicians who may not share similar goals with the hospitals. In the end, statistical analysis did not demonstrate that bundle compliance correlated directly to a reduction in specific adversee events, despite an overall reduction in AOI adverse events in the PPSI. In 2013, IHI indicatedd the Augmentation, Induction, and Vacuum Extraction bundles were being retired or modified. Each bundle was measured monthly for compliance. Specific components of each bundle whichh lagged in performance are seen in the following charts with solutions provided to help guide future improvement efforts. Induction Bundle Resultss The Inductionn bundle achieved the highest level of compliance overall, improving from a baseline of 38 percent to 88 percent, a 132 percent improvement (See Figure 4 below). To obtain higher scores, strategies to raise compliance on the tachysystole element ( light blue) were created along with other processess to support compliance with the bundle overall. These included: Creation of one standardized order set for inductions and one standardized oxytocin IV mixture. One standardized tachysystole guideline, protocol or algorithm for all stafff to follow. Implementation of prompts and reminders for staff regarding patient exams and checks prior to administering oxytocin. One team developed a tag with these prompts to be placed directly on the IV bag containing oxytocin. Creations of a policy for preventing elective inductions prior to 39 weeks gestation unless medical exception (reviewed in peer review setting). One standardized process for scheduling inductions, owned by one person per shift, who consistently applies the policy preventing electivee inductions prior to 39 weeks gestation. Figure 4 Premier Perinatal Safety Initiative 9

10 Augmentation bundle results Results for the Augmentation bundle demonstrated the second highest compliance rate, improving from 18 percent to 86 percent, a 377 percent increase. Figure 5 below demonstrates successful hospitals needed to create strategies to obtain estimated fetal weight from providers, which had the lowest overall compliance (light green line) in this bundle. The estimated fetal weight is not required to be documented as a specific weight, and some teams determined the use of LGA (large for gestational age), AGA (average for gestational age), or SGA (small for gestational age) was acceptable to comply with this element of the Augmentation bundle. Delivering practitioners noted thatt estimating specificc fetal weight is i difficult to achieve, and were concerned thatt documentation would increasee the risk of criticism of care in the event of an adverse outcome. These practitioners were counseled that there was greater risk in not documenting their consideration of fetal size when developing a labor plan, which improved individual provider compliance with this element. Vacuum bundle results Improvement on compliance with this bundle increased overall from 7 percent to 70 percent, a 900 percent increase. However, the lower final achievement level may be the result of launching the Vacuum bundle a few months after the other bundles. This bundle required more documentation and intense communication processes which lowered overall results if not documented properly. Interestingly, improvement in this bundle corresponded statistically to improvement in providing VBAC services, although increasing VBAC rates was not a focus of the project. Teams with greatest improvement in compliance with the Vacuum bundle implemented team huddles before vacuum deliveries to discuss the plan, and to establish a shared mental model. Post-delivery debriefs were also used to identify areas for improvement, processes that were effective, and to standardize documentation. This post-delivery debrief ensured the timing of the application of the vacuum devicee and the number of pop-offs are consistently and accurately recorded. Some teams provided reminder cards (or EMR prompts) for physicians to use when dictating an operative vaginall delivery to ensure all bundle elements were documented. The standardization of the estimated fetal weight field in the medical record was also critical in improving both compliance and the ability of the chart reviewer to locate the documentation to give the team credit for this element of the bundle. Figure 5 Figuree 6 10 Premier Perinatal Safety Initiative

11 Enhanced communications and teamwork Failures in teamwork and communication account for 70 percent of sentinel events in Obstetrics. 27 Recognizing this, The Joint Commission, ACOG, and the Institute of Medicine have acknowledged teamwork and communication are a critical element of patient safety. 28,29 Improved teamwork has been shown to reduce preventable adverse events, 30 such as uterine rupture and neonatal death. The PPSI was designed to increase teamwork and effective communication among perinatal teams. To be prepared to take appropriate action during worst-case scenarios, especially during emergency situations and patient hand- offs between team members, participating hospitals implemented proven strategies and conducted simulations for certain high-risk protocols. These strategies and tools included TeamSTEPPS, the Situation Background Assessment Recommendation (SBAR) and use of mannequins in simulation exercises. TeamSTEPPS is a teamwork system designed to improve the quality, safety and efficiency of healthcare. 31 Originally developed by the Departmentt of Defense in collaboration with AHRQ, the goal of TeamSTEPPS is to produce highly effective medical teams that optimize the use of information, peoplee and resources to achievee the best clinical outcomes for patients. This system can be implemented in various healthcare settingss to improve communicatio on and other crucial teamwork skills among healthcare professionals. Designed initially for the military, the SBAR standard communication tool is an effective situational briefing strategy. Using SBAR, team members communicate relevant case facts in a respectful, focused and effective manner. SBAR is useful during nurse-to- hand-offs, such as shift changes and urgent situations. Obstetric emergency simulation Simulation training for obstetric procedures physician communication or during has been shown to improve provider skills in the management of both shoulder dystocia and postpartum hemorrhage. 32 Participating hospitals used simulation exercises, featuring actresses and mannequins, to increase their teamwork and communication skills, and to practice using thesee skills during perinatal crisis situations. The teams practiced responding to perinatal emergencies, ranging from maternal hemorrhage to life-threatening infant distress. Debriefing with the teams provided critical information to guide further care and communications, and to improve the ability of the providers to act as a synchronized team. Two-and-a-halincluding didactic information of culture of days of on-site training, safetyy principles, TeamSTEPPSS curriculum, implementation strategies, and three multidisciplinary in-situ simulations (using a train the trainer model) were provided to participating hospitals. Figure 7 Premier Perinatal Safety Initiative 11

12 High Reliability Perinatal Safety Assessments PIMS performed baseline on-site High Reliability Perinatal Safety Assessments, or HRPSA, at each hospital to measure how well hospitals were meeting labor and delivery practices recommended by professional organizations and published best practices. The scores were benchmarked against all hospitals in the PPSI, and to the comparator group of hospitals. A final assessment was performed at the end of the PPSI to evaluate improvement toward becoming a high reliability organization, including adoption of standardized communicationn and simulation. Figure 8 below shows PPSI hospitals improved substantially from a baseline average score of 66 percent to an overall final average of 89 percent- status (blue) overall. (This was a 35 percent improvement overall). In addition, all 14 PPSI just one point short of achieving high reliability participants surpassed the industry standards compliance level (green). The final results were consistent with the hospitals in thee comparator group, which followed various other quality improvement methodologies (See examples in Appendixx pagee 22). This may demonstrate that an overt commitment to quality improvement and team communication may be equally impactful as the specific methodologies utilized in the effort. Teamwork and communication may help improve outcomes even when hospitals are utilizing different processes to improve their caree delivery. Bestt practices found during HRPSA included: Implementation of a birth injury improvement project to immediately spread the PPSI improvements across the entire health system. This included standardizing policies and processes, and using in-situ simulation which ledd to communication and culture changes. Highly engaged and supportive leadership and very robust IT/informatics support for quality reports. Physician champion and nurse leader review of all charts for bundle deficiency or AOI elements; investigate issues; and to provide counseling forr clinical behavioral changes. Physician issues advance to peer review if informal counseling is ineffective. Figure 8 12 Premier Perinatal Safety Initiative

13 Metrics for measuring harm reduction The PPSI hospitalss tracked dataa demonstrating several harm metrics. Ten harm measures (maternal and neonatal) made up the AOI scale (See Table 2 in the Appendix). Related metrics to AOI included the Weighted Adverse Outcome Score and the Severity Index. Birth Trauma, as defined by PSI 17 (modified) reporting criteria, was also evaluated during the project. By the end of phase 1, all 14 hospital teams were below the 2008 AHRQ provider rate for this indicator, a national 33 comparative rate measuring perinatal harm. Events of hypoxia and asphyxiaa were also tracked, as these types of injuries typically createe the most significant injuries to the neonate and highest professional liability exposures. Harm reductions PPSI hospitals reduced the annual AOI rate, which measures the number of mothers and neonates with one or more of the ten identified adverse events as a proportion of total deliveries (See Figure 9 below). The final average improvement for reduction of harms on the AOI scale compared to 2006/07 baseline scores were: 8.8% improvement for Academics 18.3% improvement Non-academics 14.4% improvement avg. overall 10 off 14 hospitals improved Figure 9 below shows scores by the hospitals in all three metrics discussedd above (improvement is green). Note that several hospitals which did not improve (red) still had scores well within the overall average range. Figure 9 Premier Perinatal Safety Initiative 13

14 Figure 10 Figure 11 AOI harm reduction Figure 10 above depicts the rate of the average number of all 10 AOI adverse events (six maternal and four infant measures) over the course of the two-year baseline and through the five years of the project. Overall, the participating hospitals averaged 58 harms per 1,,000 births at the outset, improving to a lower level of 47 per 1,000 births (a 19% improvement). There are two areas of special cause (circled) which reflect statistically significant processs shifts resulting in fewer adverse events. The first special cause is the dramatic drop whichh occurred during and immediately after the initial education concerning the clinical changes and improved communication techniques following the kick-off of the project in The second area of special cause, reflecting statistically significant lowering of the average adverse event rate, is seen over the last two years of the project. During this time the teams were adding a focus on developing expert teams through use of simulation and TeamSTEPPS principles. Thesee shifts could indicate that the improved focus of team members on the project through meetings and team communication may be as important to final results as the clinical interventions and process changes selected to improve clinical outcomes. Maternal adverse injuries decreased The PPSI hospitalss were successful in reducing the number of adverse events to mothers per 1,000 deliveries, as indicated in Figuree 11 above. Harms were reduced 20% overall from 41.5 per 1000 deliveries during baseline years of to per 1000 deliveries by the final quarter of Maternal adverse events were typically less severee than injuries to the Neonate, cost lesss to settle in litigation, and were closed in litigation faster than cases focused on neonatal harm. Some harms, such as maternal death, are very rare events and, therefore, difficult to reducee significantly or evaluate with statistical rigor. During baseline (06-07) there were eight maternal deaths reported, while there were fewer maternal deaths (six) during the longerr intervention phase (08-12). From these 14 deaths, eight malpractice cases were filed. Overall, the combined rate of maternal and neonatal deaths was reduced from baseline (.030) to end (.024)) for an improvement of 20% inn this small data set. Four hospitals (one academic and three non-academics) had no death outcomes reported during the PPSI. Although the number of neonatal deaths increased from fourr in the baseline ( ) to eight during the intervention phases ( ), the result was a slight improvement due too the significantly higher number of births in the interventions period. 14 Premier Perinatal Safety Initiative

15 Figure 12 Reduction of adverse events and claims Figure 12 above demonstrates the PPSI hospitals were successful in reducing the rate of adverse events and claims filed against them as compared to total deliveries (one hospital is omitted which had no claims filed during PPSI), as well as lowering the number of claims as part of total adverse events. The red print at lower left indicates a baseline rate of 1 claim filed per 4672 deliveries (.00021), improving during phases 1 & 2 to 1 claim filed per 5,739 deliveries (.00017), an improvement of approximately 19 percent in reducing claims per total deliveries. The results of less than two claims per 10,000 births is better than similar project results with between , 13 claims per 10,000. The blue print at lower right indicates a baseline rate of claims to adverse events (AOI)) at 1 claim filed per 260 AE s (.0039) improving to 1 claim filed per 284 AE s (.0035), an 11 percent improvementt in fewer claims filed in relation to AE s reported. While PPSI hospitals reduced both harms and claims overall, there are many demographic, personal, financial and societal influences on whether someone will file a malpractice claim. In addition, not all claims filed have legal merit, resulting in approximately 59 percent of claims being closed without loss payment. Therefore, while total claims filed against deliveries were reduced, we cannot be certain of the reasons for patients to either file or not file malpractice claims. Premier Perinatal Safety Initiative 15

16 Figure 13 Reduction in hypoxia and asphyxia harms Birth brain injuries create the highest liability exposure because the child often needs a lifetime of medical care, suffers lost wages, and has significant pain and suffering damages. Figure 13 above shows that prevention of just four birth brain injury claims by improving care could save nearly $5 million in liability costs based on the average claim losses paid by PPSI hospitals, in addition to saving countless emotional and financial hardships involved. We estimated PPSI prevented 167 inborns (infants born inside the hospital) from hypoxic events during the course of the project. Eleven medical malpractice claims for brain injury were filed from 4555 hypoxic events reported (2.4 percent filing rate) in PPSI hospitals for deliveries between The rate of filing these claims from total PPSI deliveries was The average costss of brain injury/hypoxia claims (15 closed claims in i PPSI and the AEIX comparisons) were: $2.709 million average loss paid on 7 claims actually settled, and $1.083 million average per case on 15 closed brain injury cases Effectiveness of claim processes/litigation PIMSS found that 12 of 18 (67%) hospitals interviewed had early disclosure/apology programs of whichh 85 percent of those said they include Obstetric adverse events in such programs. However, 33% said resolving Obstetric injuries was not as effective as other typess of injuries, while 58% said it was about the same and 8% found them easier to resolve in such a program. Smaller claims (loss paid $100,000 or less) were most likely to settle in 365 days or less, but only 11 of 68 (16%) of claims against PPSI and comparator hospitals weree settled on this basis. Seven (63%) of these smaller losses had no defense costs, indicating hospitals are able to settle smaller injury claims quickly and efficiently. Overall, the PPSI hospitals took an average of 453 days to settle their claims of $100,000 or less, and the AEIX comparator hospitals took 478. Of the cases settled within 365 days by thesee two groups combined, 55% were minor injury, 11% were medium injury, and 33% were major injuries. 16 Premier Perinatal Safety Initiative

17 Harm Metric Results Figure 14 Improvement in WAOI score The Weighted Adverse Outcomes Score (WAOS) in Figure 14 (a quarterly report monitoring quarterly progress) captures the severity of the AOI outcomes by weighting each outcome - the sum divided by total deliveries. The participating hospitals baseline WAOS average score of 1.19 was reduced (improved) to 1.11 at the end of Phase I in 2010, and furtherr reduced to 1.08 at the conclusion of the project in 2012, while the comparator group grew slightly worse from 1.09 to 1.26 during Phases I and II (-16%). The participating non- academic hospitals improved their overall average WAOS scores from baseline 1.19 to.97 at the end (18% improvement), while academic hospitals worsened slightly from 1.20 to 1.23 (-2.3%). Several best practices noted from the HRPSA weree considered as possibly driving this improvement: Interdisciplinary communications via OB Town Hall meetings and patient safety rounds Debriefings after critical events, twice daily huddles for unitt situational awareness Hand-offs accomplished using emplates for standardizing communication and documentation All OB providers required to undergo EFM training as condition of medical staff membership, and to participate in team training Neonatologists, pediatric intensivists, and a back-up OB were on the unit 24/7 with OB-dedicatedd anesthesiologists Labor processes were standardized Hard stop policy preventing non-medically strips for tachysystole and do not rely on documentation of someone else s indicated elective deliveries prior to 39 weeks gestation in effect Nursess audit the actual interpretation of the strips Premier Perinatal Safety Initiative 17

18 Liability Claim Results Figure 15 Potential for reducing liability exposure OB claims are often high severity in amount paid, but low frequency compared to volume of deliveries. Through June 2012, AEIX dataa demonstrated that while OB claims may be less than 10 percent of the total claims filed against a hospital annually, they can account for 25 percent or more of the total losses paid to patients for all claims resolved by payment. The PPSI evaluated claims dataa from 14 participating and 8 comparator hospitals, and also looked at claims data from AEIX hospitals whichh had similar demographics to the PPSI participating hospitals. Of 68 closed OB claims studiedd for the participating and comparator groups, 23 (34 percent) were maternal-only and 45 (66 percent) were neonatal (a few had secondary maternal claims). Losses were paid on 28 claims (41 percent), meaning a zero closing rate ( no losses paid) of 59 percent on OB claims filed. The average of paid-only claims was $2,230,909 each, or over the total 68 claims was $852,995 each. Losses ranged from as low as $1, 910 to $16 million. Of the 68 claims, 13 percent paid $1million or more. Losses were paid on 19 of the neonatal claims (42 percent), with an average per case of $3,044,099 (165 times higher than maternal loss average). The average losss over all 45 neonate cases was $1,285,286 per case filed. Defense expensess over the 45 cases averaged $95,008 per case.. Approximately 59 percent of all OB claims were closed withoutt loss payment. PPSII hospitals reduced liability exposuree In addition to the savings on hypoxia losses, the PPSII hospitals reduced their overall number of new claims being filed on deliveries during PPSI compared to deliveries during baseline (Seee Figure 15 above). Of those new claims, more were being resolved without loss payout than during the baseline period. Figure 15 ends in 2011 to allow time for closed claims to develop and allowing two years filing lag time. The delays in filing and resolution of claims ( lag time ) can impair accuracy of claim data. Most OB claims are filed within two years of birth, but many states allow filing claims to age 18. It may take several years for OB claims to resolve, especially whenn loss is paid. In PPSI, approximately 92 percent of historical OB claims had been filed withinn two years of injury. This compares favorably to other industry reports of 89.6 percent filed in that time The average lag time from notice of claim to resolution during PPSI was months when claims settled under $100,000, and 30 months whenn higher losses were paid. The PPSI results were better than reported in another study which foundd 95 percent of their casess resolved at 36 months, while OB cases required 60 months to resolve by their hospitals due to high severity of payouts Premier Perinatal Safety Initiative

19 Liability Claim Results Figure 16 Reducing liability exposure Another way of examining financial benefit of reducing perinatal adverse events is looking at the financial estimates of the hospitals for their claim exposures in the baselinee period, compared to their estimates of exposure for those claims arising from births during the intervention period (See Figure 16 above). The reserves set aside to pay OB cases, along with payments already made ( total incurred ) by the PPSI participating hospitalss (black), is compared to the similar group of AEIX hospitals (red). The incurred total for participating hospitals (See red circles above) was reduced from $17.98 million for births with injuries arising in the baseline period, compared to only $4.65 million for births with injuries in the Phases period (74 percent reduction). In contrast, the (red) AEIX hospitals not participating in PPSI increased their financial exposure during the same timeframe from $12.57 million to $17.86 million (42 percent increase). Defense cost averages were similar for both groups, with each improving against baseline. This may be due to lower claim frequency reported in the industry from 2004 through Severity over the same time has increased in the industry 36 demonstrating the reduction in loss severity by PPSI hospitals is an important result. Finding and resolving OB claims earlier PPSII evaluated what type of perinatal injuries moree often resulted in a lawsuit filed, and whether UB04 and ICD-9 hospital billing data mightt raise red flags to earlierr identify cases likelyy to result in claims. Unfortunately, when comparing claims actually filed to data entered at thee time of birth, there were no patterns demonstrating which adverse events were moree likely to become claims. Numerous claims with serious injuries did not have hospital coding of harm, indicating injuries were either not detectable at birth, arose later, or simply were nott captured by coding at birth. Somee generalities were noted which could help providers find events which might later become claims, so that earlier intervention and resolution by the hospital becomes possible. Claims were more likely to be filed whenn injuries weree noticeable and were perceived as not intended by, or related to, the proposed labor and delivery treatment. Serious injuries (life altering with long term care required) comprised a higher rate of claims filed (61 percent) than medium injuries (6 percent) and minor injuries (17 percent), whilee 6 percent had unknown injuries. The minor injuries were typically lacerations or allergic reactions which were noticeable, but whichh caused only small injuries. Premier Perinatal Safety Initiative 19

20 Conclusions CONCLUSIONS The 14 PPSI participating hospitals over five years reduced adverse events and patient harms, achieved a reduction in malpractice claims and associated liability costs, and significantly improved provision of clinical care bundles to patients. In addition, some improvements occurred which were not the focus of the clinical interventions, such as lowering the primary cesarean section rates and increasing rate of VBAC. These additional improvements may be a result of the attention to improving communication and teamwork, as would be expected based on prior studies in this area. Successful hospitals overcame disruptions caused by turnover in clinical staff involved in managing the quality project, and which leveraged their physician champions when outside healthcaree providers were hesitant to comply with process changes. Significant improvement in high reliability performance was noted, as scores from on-site assessments (HRPSA) rose betweenn baseline and project end, with the overall project average falling only one point short of the highest rating category. In addition to ensuring consistency in newly created clinical care processes, hospitals desiring to succeed in lowering patient harms also needd to provide staff education and assesss provider technical capability as part of ongoing priorities. The analysis of malpractice claim files indicated independent decision making and communicationn were keys to achieving better outcomes, as process improvement measures alone didd not eliminate many of the errors which caused injuries which resulted in lawsuits. Obstetric harms to patients, though low in frequency, often create significant physical and financial harm to patients and hospitals. Lowering patient harms, therefore, should be a priority in Obstetrics as even small improvements can provide significant benefits. Hospitals whichh can demonstrate lower rates of adverse eventss and less liability exposuree will save money for their healthcare systems, and will be attractive to health insurers, liability insurers, and patients. 20 Premier Perinatal Safety Initiative

21 Appendix APPENDIX Background PPSI project design and assessment goals were set in November 2007, after involving leadership representatives from IHI, AHRQ, the National Perinatal Information Center (NPIC) and additional national obstetrics-related professional organizations. A pre-project onsite High Reliability Perinatal Safety Assessment (HRPSA) was conducted by PIMS for each participating hospital. Participants were provided with a baseline report against which they could monitor and track performance. Monthly team conference calls, quarterly webinars and access too a Premier perinatal web portal allowed participants to view current topics, share best practices, update team data and view team success. A second follow up HRPSA was performedd in 2012/13, which allowed teams to measure their progresss in the project and prioritize further improvement goals. To monitor progress toward the PPSI goals, hospital progress reports on the reductionn of harm were provided to teams and hospital leadership on a quarterly basis. To prevent a decline in patient safety improvement efforts and support the continued engagement t of the healthcare team, participant hospitals shared the results of the monthly bundle chart audits with the entire perinatall unit staff and delivering practitioners. Premier Perinatal Safety Initiative 21

22 Appendix Interventions and otherr quality projects The chart above indicatess the specificc interventions provided forr all PPSI hospitals. Many of the hospitals in both the PPSI and comparator group had other quality projects ongoing as part of labor and delivery services, and some are indicated below: 22 Premier Perinatal Safety Initiative

23 Appendix High Reliability Characteristics often cited for high reliability perinatal services include*: An organizational culture where patient safety is promoted, supported and understood throughout the organization Strong interdisciplinary leadership Professional team interaction thatt promotes the communication of important patient information and expedites the prompt delivery of medical attention during emergencies Multidisciplinary rehearsal of emergencies Adoption of common language to describe fetal well-being during labor amongst healthcare providers and nurses Policies and procedures which are supported by national professional standards, evidence-based medicine and best practices supporting patient safety Standardization and simplification of clinical protocols and unit operations The High Reliability Characteristics categories that are assessed and scored in the HRPSA are: Team work and communication Reliability concepts Situational awareness Fetal monitoring Care bundles Critical event management Risk & liability exposure *Sources: G. Eric Knox, MD, Kathleen Rice Simpson, Ph.D., RNC, FAAN and Kathryn Eblen Townsend, JD, RN, ARM, High Reliability Perinatal Units: Further Observations and A Suggested Plan for Action, ASHRM Journal, Fall 2003, pg Barbara J. Youngberg, BSN, MSW, JD, FASHRM, Assessing Your Organization s Potential to Become a High Reliability Organization, ASHRM Journal 2004, Vol. 24,No.3, pg ) Report of findings and benchmarking from the HRPSAA on-site assessments helped hospitals measure progress toward becoming high reliability in labor and delivery: With the HRPSA narrative report, each hospital received a table displaying the criteria evaluated and scored for each characteristic, and benchmarked against the results of Premier s Perinatal Safety Initiative hospitals, along with pertinent recommendations. The criteria used to evaluate each hospital s progress weree derived from certain industry standards of several professional organizations including, but not limited to, ACOG, AWHONN, AAP, AORN, ASA, and certain accreditation organizations (e.g. The Joint Commission), and regulatory agencies (e.g. CMS, FDA, DEA, NPDB). Premier Perinatal Safety Initiative 23

24 Appendix Variables and Measures Input measures PPSI participants used two tools to gather data on input measures: the HRPSA and the Hospital Survey of Culture of Safety (HSOPSC) developed by AHRQ. PIMS developed the HRPSA for the collaborative based on evidence-based clinical practices, the most current standards of practice set forth by the Joint Commission, guidance from a variety of physician and nurse professional organizations, and high reliability standards from other published works on this topic. The HRPSA consists of 48 items regarding perinatal safety policies and practices needed to ensure a high reliability environment for perinatal safety. The assessment team gathered information to determine which patient safety practices the hospital and the perinatal unit had in place, and whether those practices were consistentt with characteristics s often observed in high reliability perinatal services. These characteristics include: An organizational culture where patient safety is promoted, supported and understood throughout the organization Strong interdisciplinary leadership Professional team interaction that promotes the communicatio on of important patient information and expedites the prompt delivery of medical attention during emergencies Multidisciplinar ry rehearsal of emergencies Adoption of common language to describe fetal well-being during labor among all healthcare providers Policies and procedures which are supported by national professional standards, evidence-based medicine and best practices supporting patient safety Standardization and simplification of clinical protocols and unit operations To measure subjective impressions of the culture of safety att each intervention hospital, obstetricians, pediatricians, anesthesiologists, nurses and ancillary staff at each hospital were asked to complete the HSOPSC beforee the project launched and the onsite reliability assessment was performed. These results established their baseline culture of safety measure. A two-year follow-up survey was administered at each hospital in 2010, followed by a third follow-up survey in The HSOPSC reports the organizations safetyy culture using twelve dimensions: 1. Teamwork within units 2. Supervisor/manager expectations and actions promoting patient safety 3. Organizational learning continuous improvement 4. Management support for patient safety 5. Overall perception of patient safety 6. Feedback and communication about error 7. Communication openness 8. Frequency of events reported 9. Teamwork across units 10. Staffing 11. Handoffs and transitions 12. Non-punitive response to errors Whilee slight improvement was made on HSOPSC scores, the improvement on culture of safety was not statistically significant. This is nott surprising, since the Agency for Healthcare Research and Quality Safety Culture database has only demonstrated a 1 percent aggregatee improvement across the U.S. during the past six years. Best practices to improve culture of safety remain somewhat elusive. 24 Premier Perinatal Safety Initiative