If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Begun in 2003, the Yale-New Haven Hospital comprehensive obstetric safety program consisted of measures to standardize care, improve teamwork and communication, and optimize oversight and quality review. Prior publications have demonstrated improvements in adverse outcomes and safety culture associated with this program. In this analysis, we aimed to assess the impact of this program on liability claims and payments at a single institution. We reviewed liability claims at a single, tertiary-care, teaching hospital for two 5-year periods (1998-2002 and 2003-2007), before and after implementing the safety program. Connecticut statute of limitations for professional malpractice is 36 months from injury. Claims/events were classified by event-year and payments were adjusted for inflation. We analyzed data for trends as well as differences between periods before and after implementation. Forty-four claims were filed during the 10-year study period. Annual cases per 1000 deliveries decreased significantly over the study period (P < .01). Claims (30 vs 14) and payments ($50.7 million vs $2.9 million) decreased in the 5-years after program inception. Compared with before program inception, median annual claims dropped from 1.31 to 0.64 (P = .02), and median annual payments per 1000 deliveries decreased from $1,141,638 to $63,470 (P < .01). Even estimating the monetary awards for the 2 remaining open cases using the median payments for the surrounding 5 years, a reduction in the median monetary amount per case resulting in payment to the claimant was also statistically significant ($632,262 vs $216,815, P = .046). In contrast, the Connecticut insurance market experienced a stable number of claims and markedly increased cost per claim during the same period. We conclude that an obstetric safety initiative can improve liability claims exposure and reduce liability payments.
The health care safety and quality movement has multiple goals, including (1) improvement of quality of care for individual patients, (2) reduction in the incidence of and exposure to adverse events, and (3) control of health care spending through accountable and value-based care. Preventable medical errors and mishaps diminish the ability to achieve all 3 goals, and thus efforts to control their occurrence are taking center-stage in health care improvement discussions.
Patient safety interventions have demonstrated remarkable improvements in quality indicators and reductions in adverse outcomes. However, less is known about how such interventions impact health care costs. Reducing waste and the spending required to respond to adverse outcomes is one way to reduce costs. It is also presumed that improvements in safety culture and the resultant enhanced collaboration and teamwork results in staffing efficiencies, such as less staff turnover and fewer staff vacancies. Finally, quality improvement efforts may alleviate some medicolegally-motivated defensive medicine practices complicating health care.
The contribution of medicolegal concerns to direct and indirect health care costs is a subject of debate. However, with obstetrics in a chronic professional liability insurance crisis, and with liability insurance and defense consuming a considerable amount of financial resources in obstetrics, demonstrating an impact on medicolegal outcomes, in addition to adverse outcomes, is an important goal in this field. Fewer lawsuits may be a surrogate marker of improved outcomes, but are probably a valuable indicator on their own. Decreasing claims also would reduce the overhead costs associated with legal defense and should also reduce overall payments for awards and settlements.
In 2002, Yale-New Haven Hospital (YNHH) partnered with its liability insurance carrier (MCIC Vermont, Inc., New York, NY) to introduce a comprehensive obstetrics safety initiative aimed at improving quality of care and reducing liability costs. We have previously demonstrated reductions in adverse outcomes and improvements in safety culture/climate associated with this program.
More than 3 years after the maturity of this program, we now aim to describe the changes in our liability profile, namely the number of and payments for obstetric legal cases.
Materials and methods
We incrementally introduced multiple patient safety interventions from Dec. 2002 to Nov. 2006 at a university-based obstetrics service at YNHH. The details of this program have been previously described.
Briefly, the core elements of this project included:
Outside Expert Review: we began in 2002 with a review of our obstetric services by 2 independent consultants. This site visit culminated in recommendations that focused on principles of patient safety, evidence based practice, and consistency with standards of professional and regulatory bodies.
Protocols and Guidelines: protocol and guideline development began in 2004 with the aim to codify and standardize existing practices. Over 40 documents were produced during the study period.
Obstetric Safety Nurse: an obstetric safety nurse was hired in 2004 to facilitate planned interventions and assist in data collection.
This nurse was in charge of educational efforts—including team training and electronic fetal heart rate (FHR) monitoring certification—and operations relating to patient safety activities.
Anonymous Event Reporting: we initiated in July 2004 a computerized and anonymous event reporting tool (Peminic Inc, Princeton, NJ) that allows any member of the hospital to report an event or condition leading to harm (or potential harm) to a patient or visitor. Reports were reviewed and investigated.
Obstetric Hospitalists: resident supervision and leadership of the inpatient activities was assumed by our Maternal-Fetal Medicine team to provide 24-hour, 7-day a week in-house coverage, beginning in 2003.
Obstetric Patient Safety Committee: established in 2004 this multidisciplinary committee of physicians, midwives, nurses, and administrators provides quality assurance and improvement oversight. In particular, this group met monthly to review adverse events and address the needs for protocols and policies.
Safety Attitude Questionnaire: to assess employee perception of teamwork and safety, we annually surveyed our teams with this tool, adapted from the aviation field.
Team Training: we implemented crew resource management seminars, based on those of airline and defense industries. These 4-hour classes included videos, lectures, and role-playing with the goal of integrating obstetric staffing silos (physicians, midwives, nurses, administrators, assistants) and teaching effective communication. Completion of the seminar was a condition for employment and/or clinical privileges.
Electronic FHR Certification: teaching for this included dissemination and review of NICHD guidelines, review of tracings, allocation of study guides, and voluntary review sessions, culminating in a standardized, certified examination. All medical staff and employees responsible for FHR monitoring interpretation were obligated to pass this exam at program inception or within 1 year of employment.
Events, claims, and suits related to obstetric cases at YNHH were collected prospectively by the liability carrier (MCIC Vermont, Inc.) for the hospital and all of its employees and providers, and classified according to event year. MCIC Vermont, Inc. covers all care at YNHH, including professional liability insurance for all obstetricians and midwives. For the purposes of this study, only formal claims and suits filed against the hospital or a hospital provider were designated as ‘cases.’ A case consisted of a claim or suit requesting financial compensation of the patient for alleged harm and resulting in legal involvement and/or response by the liability carrier.
This includes cases dropped by the plaintiff or settled with or without payment before the filing of a formal lawsuit. Events noted by the legal or medical liability teams to be at risk for legal action were not included.
Cases were categorized according to high, moderate, or low severity, as described in Table 1, by the liability carrier using the industry standard National Association of Insurance Commissioners Index.
Closed cases were defined as those resolved by withdrawal, court judgment, or settlement. Open cases were claims or suits filed in court but still unresolved at the time of performing the analysis. Connecticut state law (CGS § 52-584) requires that a medical malpractice lawsuit must be initiated within 2 years from the date the injury is first sustained or discovered (statute of limitations), or 3 years from the date of the act or omission causing the injury (statute of repose).
Thus, a malpractice claim must be initiated within 3 years of the act/omission even if the injury is not discovered until after 3 years have passed. There is no law extending the statute of limitations for injured minors. Thus, obstetric cases up to Dec. 2007 must have been filed before Jan. 2011, ensuring complete accounting for all possible cases in this study. Study completion date of Dec. 2007 was chosen to allow for the statute of repose as well as a subsequent 18-month period to allow any open cases to resolve.
Indemnity payments were identified by our liability carrier and include all compensation to claimants of plaintiffs. Payments do not include costs of investigating or defending the case or other allocated loss adjustment expenses. As events that did not lead to claims or suits were not included, dollars held in reserve for possible future actions were not included. All monetary values are expressed in dollars and adjusted for inflation to reflect 2007 values, according to the Consumer Price Index.
There were no concurrent changes in malpractice law on caps or noneconomic damages in Connecticut during this study period. A statute requiring a ‘certificate of merit’ from a qualified health care provider for medical liability cases was passed in 2005 (CGS § 52-184c and 52-190a). There were no institutional changes in mediation or adverse event disclosure policies during the study period.
Analysis was performed tracking the number of liability cases per 1000 deliveries, per year. Cases were normalized per 1000 deliveries to control for any variation in volume across study years or periods. Comparisons were made for 2 5-year periods (before study inception [Jan. 1998-Dec. 2002] and after study inception [Jan. 2003-Dec. 2007]) using Student's t test, the median test, Mann-Whitney U test and χ2 or Fisher exact test where appropriate. Poisson regression was used to analyze annual trends in numbers of claims per 1000 deliveries. In addition, analysis of differences and trends in annual liability payments was performed on closed as well as open and closed (combined) cases. For combined case payment analysis, we used the overall median liability payment for the 5 surrounding years as the estimate for each open claim, assuming each open case resulted in payment. Cases that did not result in payment were not included in payment analyses. We performed the additional analysis of combined cases because a closed claim analysis may bias results in favor of the second epoch, given that it is likely to have more open claims. When claims remained open we performed worst-case and best-case scenario analyses when estimating the numbers of claims settled without payment. Worst-case scenarios designated open cases as being settled with payment, whereas best-case scenarios designated them as settled without payment. P values < .05 were considered statistically significant. Analysis was performed using commercially available software (SPSS version 18.0; SPSS, Inc., Chicago IL).
This project was reviewed by the Chair of the Yale University Human Investigations Committee and was deemed a quality assurance activity and thus not required to undergo review by the Committee.
Our unit averaged approximately 4600 deliveries annually, with no statistically significant difference between both epochs (Table 2). We identified 44 cases overall during the entire 10 year study period, with 30 of those associated with events before initiation of our safety initiative and 14 after. Twelve (12) cases resulted in no payment made, with 7 of these in the 5 years before our patient safety project and 5 cases after the initiation of our intervention (Table 3). There were 2 open claims remaining at the time of this report, both being in the second 5-year epoch.
Table 2Comparison of outcomes before and after program inception
Annual cases per 1000 deliveries decreased significantly over the study period (Poisson regression, P < .01; Figure 1). Compared with the rates before initiation of our program, median annual rates of cases per 1000 deliveries were significantly lower after study inception (1.31 before vs 0.64 after, P = .02; Table 2 and Figure 1). Distribution of cases by severity and distribution of cases by type, however, did not significantly change after inception of our patient safety program (Table 2). The number of cases resolved without payment did not significantly change, both in the closed case analysis (n = 7 [23%] vs n = 5 [42%]; P = .27) and in worst-case and best-case scenarios in the combined case analysis (worst-case: n = 7 [23%] vs n = 5 [35%]; P = .48; best-case n = 7 [23%] vs n = 7 [50%]; P = .19).
Closed-case analysis revealed that payments were drastically reduced after the patient safety effort, from $50.7 million to $2.2 million (Table 2). Median annual payments, per 1000 deliveries, were significantly lower in the second time period as well ($1,141,638 vs $63,470; P < .01); this statistically significant result held true when performing the combined (open and closed) case analysis as well (Table 2). However, annual trends towards reduced payments, both in the closed case and combined case analyses, were not statistically significant. Figure 2 represents a graphic depiction of the yearly trend for the combined case analysis; the closed case analysis does not appear different.
To determine whether the patient safety program had any impact on payments to claimants, we analyzed how payments differed across both time periods. The median monetary amount per case resulting in payment to the claimant was statistically significantly different in the combined case analysis ($632,262 vs $216,815; P = .046) and in the closed case analysis ($632,262 vs 81,714; P = .03). Furthermore, there was much less variability in payments, as reflected in a narrowing of the interquartile ranges after initiating our safety program (interquartile range before $2,996,068, vs after $270,361 [combined cases] and $267,280 [closed cases]).
This analysis demonstrates a strong association between introduction of a comprehensive obstetric patient safety initiative and a dramatic reduction in liability claims and liability payments. We have estimated a 95% reduction in direct liability payments and a savings of $48.5 million over a 5-year period. We also see a consistent pattern of statistically significant trends in reduced payments and in the variability of these payments. Furthermore, during this patient safety intervention there was a 53% reduction in liability claims and lawsuits compared with the 5 years prior. The mean number of annual cases consistently dropped over the 10-year period. We were unable to see differences in the distribution in the quality (severity and types) of the cases, which may be due to small sample sizes, though there were absolute decreases in each category.
There are several limitations to this study. It is important to note that our 2 remaining open claims are in the second study period, and this may bias the results toward showing a difference between the 2 study periods when there is not one in reality (ß-error). Increasing time from injury to case closure (the ‘age of the claim’) is also typically associated with a larger final payment. However, there is not an association of age of claim and whether any payment at all is made. In Connecticut, approximately 50% of malpractice claims result in payment and there is no association with the age of the claim.
As a result, nonpayment for either claim still open in our study would strengthen our results. We believe that our estimate for this report is fair, and that the timely reporting of these results (ie, not waiting until all cases have been finalized, which on average in Connecticut is 5 years after the date of injury) is important for the obstetrics, medicolegal, and patient safety communities.
Our study is also limited by an inability to directly compare with a control group. In our case, we chose the time period before our safety initiative as a comparison. Our institution overall did not experience a statistically significant reduction in claims in nonobstetric fields (eg, surgery, emergency department, medicine, etc), when comparing the same 2 epochs (P = .16), suggesting that this was a change specific to our program rather than a generalized institutional phenomenon. Controls outside of our institution would be difficult to find and/are problematic. First, there is the issue of reporting; institutions are generally very guarded with respect to reporting their liability experiences to outside entities. To put our report into context, however, the Connecticut State Insurance Commissioner has reported that from 2005-2009 the values of claims either awarded or settled actually increased.
In terms of claim numbers, closed claim data from the Connecticut State Insurance Commissioner has reported that the total number of medical liability claims in Connecticut closing in 2010 (693) was only negligibly different from those closing in 2006 (714); more discrete data such as those focused on obstetric claims or those sorted by event year are not available.
Though not definitive proof, these data suggest that the certificate of merit statute passed in our state had little effect on numbers of claims submitted by plaintiffs. Comparisons to institutions outside of Connecticut are also limited, as other states will have different malpractice environments and few have statutes with such short conditions of repose. However, national rates of claims, as well as the severity of claims, have been reported as increasing, with obstetrics playing a key hospital risk area in this rise.
A major strength of this paper lies in its analysis by event year, rather than policy year. Although neither method allows for strict conclusions to be made about causation, analyzing by event year allows us to make stronger temporal associations. Policy year analysis would not necessarily reflect adverse events from a particular time period, as it is a measure of claims filed in a particular year without regard to when they actually occurred. This is further enhanced by Connecticut's short statute of limitations, which makes an analysis 3 years after the final claim year possible. Thus, we are able to analyze a nearly completed dataset of actual claims and payments, rather than an experience based on reserves,
The results from this analysis document a third benefit of initiating a comprehensive obstetric patient safety effort: possible cost savings. Although the primary motivations driving patient safety efforts are improving quality of care and eliminating harm, these data are also important for demonstrating further downstream impacts patient safety projects can have. The reduction in claims and payments, strictly within the context of liability concerns, saves direct legal costs, minimizes time devoted to investigation and defense, and minimizes the emotional and social costs on health care providers involved in these cases. This is particularly relevant in obstetrics, as the medical liability crisis has hit obstetrics particularly hard. The 2009 American College of Obstetricians and Gynecologists “Survey on Professional Liability” reported that 90.5% of respondents indicated they experienced at least 1 liability claim during their careers, with an average of 2.69 claims per physician.
The significance of these results outside of the narrow medicolegal context should not be underestimated. A reduction in liability claims is likely a hallmark of an environment with improved quality. In fact, coupling these results with our prior report demonstrating reduced adverse outcomes suggest a direct association, as others have reported.
Initial resistance to such programs is common, if not ubiquitous, particularly from the viewpoint that system changes seemingly act counter to individual decision-making or skill. Others have proven the value of formalizing standardization in nonacademic settings
; the findings at our site—which combines a resident service, midwifery practices, community physician practices, and a university-based maternal-fetal medicine group—can have impact in a diverse academic institution.
Furthermore, given the striking reductions in liability payments seen one cannot ignore the economic relevance of this report, particularly in today's health care environment of accountable and value-based care. Savings in legal costs beyond direct payments to plaintiffs are likely. Legal defense costs in Connecticut average from $58,000 to $70,000 per claim, including for claims that result in no payments to claimants.
A study involving a random sample of 1452 closed malpractice claims from 5 insurance carriers estimates that the administrative costs of litigating claims increases the cost of these claims by an additional 54% of the compensation paid to plaintiffs.
It is difficult to say that projects like this will have an impact on overall health care spending, however. Some experts estimate that legal fees, payments, and insurance premiums contribute to only 0.5% of US health care costs
Our study does not address the cost or efficiency of the services that were rendered over the study period. Although we did not specifically encourage any defensive practices during the study period, we did note that our cesarean delivery rate increased over time, in step with national trends.
We have no information as to whether this increase affected the risk of adverse outcomes, but we are sure that it did increase costs to patients and their insurers. Furthermore, few of these efforts can be provided at no cost, although the simplicity of many tools (such as checklists) challenges any arguments against them. Whether patient safety projects provide a net cost benefit is difficult to calculate and not known at this time. Initial costs of our program, supported by our liability carrier, are estimated at $210,000, with ongoing yearly costs of $150,000, giving a 5-year estimate of $810,000. Thus, we may estimate a substantial return on investment from the view of our medical liability carrier, on the order of 58:1.
Certainly, our effort is not the only approach to quality and safety with possible impacts on the medical liability climate in obstetrics. For instance enhanced communication skills may improve provider-patient relations after an adverse event or medical error. In fact, formal implementation of a disclosure program that also offers compensation for medical errors has shown a decrease in claims.
Others have demonstrated that most payments for obstetric malpractice cases are a result of substandard care resulting in preventable injury, adding that over 50% of litigation costs could be avoided with practices such as 24-hour obstetric coverage, adherence to medication protocols, and improved documentation, particularly in cases of shoulder dystocia.
Unfortunately, we are unable to conclude which of the core elements of our patient safety project had the most impact in achieving the results reported here.
Although improving the medical liability climate has generated much discussion, little clinically based work has actually impacted this serious problem. President Obama and the Department of Health and Human Services made patient safety projects an important part of health care reform, explicitly connecting them to improving the medical liability environment.
A first step toward this end is for the medical profession to put effective interventions in place that reduce events that result in liability. We believe this report is an important advance toward this end and is particularly important because it impacts the point of care, rather than the political or statutory structures of the medical liability machine. However, whereas we have been able to demonstrate that patient safety efforts can have a significantly positive effect on liability exposure, we do not believe that it can happen without a broad effort to improve the general liability environment. We believe a patient safety program can be even more successful in regions that have embraced meaningful tort reform, when the threat of suit is less likely to be a principle driver of the desire to reduce harm to patients.
We would like to acknowledge MCIC Vermont, Inc., its leadership, and the individual hospitals of MCIC Vermont that contributed with similar patient safety initiatives at their own institutions.
Impact of a comprehensive patient safety strategy on obstetric adverse events.