460
THE RED SECTION
nature publishing group
Effect of a Health System’s Medical Error Disclosure Program on Gastroenterology-Related Claims Rates and Costs Megan A. Adams, MD, JD1, B. Joseph Elmunzer, MD1 and James M. Scheiman, MD, FACG1 OBJECTIVES:
In 2001, the University of Michigan Health System (UMHS) implemented a novel medical error disclosure program. This study analyzes the effect of this program on gastroenterology (GI)-related claims and costs.
METHODS:
This was a review of claims in the UMHS Risk Management Database (1990–2010), naming a gastroenterologist. Claims were classified according to pre-determined categories. Claims data, including incident date, date of resolution, and total liability dollars, were reviewed. Mean total liability incurred per claim in the pre- and post-implementation eras was compared. Patient encounter data from the Division of Gastroenterology was also reviewed in order to benchmark claims data with changes in clinical volume.
RESULTS:
There were 238,911 GI encounters in the pre-implementation era and 411,944 in the postimplementation era. A total of 66 encounters resulted in claims: 38 in the pre-implementation era and 28 in the post-implementation era. Of the total number of claims, 15.2% alleged delay in diagnosis/misdiagnosis, 42.4% related to a procedure, and 42.4% involved improper management, treatment, or monitoring. The reduction in the proportion of encounters resulting in claims was statistically significant (P = 0.001), as was the reduction in time to claim resolution (1,000 vs. 460 days) (P < 0.0001). There was also a reduction in the mean total liability per claim ($167,309 pre vs. $81,107 post, 95% confidence interval: 33682.5–300936.2 pre vs. 1687.8–160526.7 post).
CONCLUSIONS:
Implementation of a novel medical error disclosure program, promoting transparency and quality improvement, not only decreased the number of GI-related claims per patient encounter, but also dramatically shortened the time to claim resolution.
Am J Gastroenterol 2014; 109:460–464; doi:10.1038/ajg.2013.375
Medical error disclosure, or “disclosure, apology and offer” (DA&O), programs have been implemented as novel strategies to decrease medical liability claims and costs, while also promoting disclosure, transparency, and institutional quality improvement (1). In 2001, the University of Michigan Health System (UMHS) implemented a DA&O program that revolutionized the claims management landscape of the Health System and has served as a nationwide model for liability reform. The details of this program have been reported previously (1,2). In brief, although before 2001 the UMHS employed a traditional “deny and defend” approach to medical malpractice claims, as of July 2001 the UMHS implemented a DA&O program designed to efficiently respond to incidents of clear medical error with open disclosure, an offer of compensation, and efforts at quality
improvement. At the same time, UMHS continued to vigorously defend itself against claims when it was determined that reasonable care was provided. A system-wide analysis in 2010 found a 36% decrease in the monthly rate of new claims per 100,000 patient encounters, 30% decrease in median time to claim resolution, and 44% decrease in average cost per lawsuit following program implementation (3). However, no discipline-specific studies have yet been undertaken to determine whether these findings are reproducible on an individual specialty level, as there are clearly wide variations in the scope of practice between specialties. Our study aimed to expand on the findings of the 2010 analysis by examining the effect of the UMHS DA&O program specifically on GI-related claims and costs.
1
Division of Gastroenterology, Department of Internal Medicine, University of Michigan Medical School, Ann Arbor, Michigan, USA. Correspondence: Megan A. Adams, MD, JD, Division of Gastroenterology, Department of Internal Medicine, University of Michigan, 3912 Taubman Center, SPC 5362, 1500 East Medical Center Drive, Ann Arbor, Michigan 48109, USA. E-mail: [email protected] The American Journal of GASTROENTEROLOGY
VOLUME 109 | APRIL 2014 www.amjgastro.com
THE RED SECTION
Methods
Following UMHS Institutional Review Board approval, we performed a retrospective review of de-identified claims data in the UMHS Risk Management Database between 1990 and 2010, naming a gastroenterologist. We hypothesized that, similar to the global analysis, the program would result in a significant reduction in time to claim resolution and costs with respect to GI-specific claims. We also explored whether this effect was specific to certain categories of claims. For the purposes of this study, a “claim” was defined as an assertion of wrongdoing that forms the basis for a request for compensation. Claims were independently divided into pre-determined categories: (i) delay in diagnosis/misdiagnosis, (ii) procedure-related (including sedation, nonsedation, and consent-related), and (iii) improper management, treatment, or monitoring. Any disagreement in categorization was resolved by consensus. Data regarding each claim, including time to claim resolution, disposition, and total settlement dollars and expenses, were reviewed. Mean total dollars incurred per claim were compared between the 10-year pre-implementation and 10-year post-implementation eras. These costs were subdivided into payments to patients (settlement and/or jury awards) and administrative/legal costs (nonsettlement-related expenses). Divisional patient encounter data were also reviewed in order to benchmark claims data with changes in clinical volume. As the number of inpatient encounters was relatively static (13,736 inpatient units of
Delayed diagnosis/ misdiagnosis of GI-related issue
15.2% 42.4%
GI procedural complications (sedation, nonsedation, and consent-related claims)
42.4% Improper management/ treatment/monitoring
Figure 1. Distribution of claims by category (1990–2010). GI, gastroenterology.
care in 2000 vs. 12,617 in 2009), a mean of the total number of procedures plus clinic visits was used to estimate clinical activity. Statistical methods Because expenditures in the two eras were not normally distributed (Skewness/Kurtosis test for normality, P < 0.001), the mean expenditures per claim were compared using the two-sample Wilcoxon rank-sum (Mann–Whitney) test. The difference in the proportion of encounters resulting in a claim in the two eras was analyzed using Fisher’s exact test (two-sided significance level 0.05). Time to claim resolution was compared using a Kaplan– Meier analysis.
Results
A total of 66 claims met inclusion criteria and were analyzed in the study. Of these, 15.2% alleged a delay in diagnosis/misdiagnosis, 42.4% related to a GI procedure, and another 42.4% involved improper management, treatment, or monitoring (Figure 1). In examining the distribution of claims by category in the two eras, several trends emerged. First, the percentage of claims alleging delayed diagnosis/misdiagnosis was lower in the postimplementation era than the pre-implementation era (21.1% vs. 7.1%, P = 0.007). Second, there were a larger percentage of procedure-related claims post-implementation (34.2% vs. 53.6%, P = 0.33; Figure 2). The breakdown of total procedure-related claims was as follows: 12 colonoscopy, 8 endoscopic retrograde cholangiopancreatography, 5 esophagogastroduodenoscopy, 2 capsule endoscopy, and 1 liver biopsy. There were 238,911 GI encounters in the pre-implementation era vs. 411,944 post-implementation (Figure 3). A total of 66 encounters resulted in claims, 38 pre-implementation and 28 post-implementation (despite a 72% increase in clinical activity), representing a significant reduction in the proportion of encounters resulting in claims (P = 0.001). The rate of claims per 1,000 patient encounters decreased from 0.160% pre-implementation to 0.068% post-implementation, as did the mean total liability per claim ($167,309 pre vs. $81,107 post, 95% confidence interval: 33682.5–300936.2 pre vs. 1687.8–160526.7 post; Table 1).
Pre-implementation
21.1%
Post-implementation
GI procedural complications
44.7%
34.4%
7.1%
Delayed diagnosis/ misdiagnosis
Improper management/ treatment/ monitoring
39.3% 53.6%
Figure 2. Distribution of claims by category in the pre-implementation vs. post-implementation eras. GI, gastroenterology.
© 2014 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
461
462
THE RED SECTION
In examining mean total liability per claim by claim type, there was a notable reduction post-implementation in both the procedure-related and improper management/treatment/monitoring categories (Figure 4). There was a marked increase in mean total liability per claim in the delayed diagnosis/misdiagnosis category due to a single outlier claim. The first quartile estimate of time to claim resolution for claims in the pre-implementation era was roughly twice that for claims filed post-implementation (1,000 vs. 460 days; P < 0.0001; Figure 5). Total settlement dollars were $4,743,726 in the pre-implementation era vs. $2,176,065 in the post-implementation era (a 54% decrease). Similarly, claims-related expenses (nonsettlement-related, including attorney’s fees and expenses from outside counsel) were $1,630,770 in the pre-implementation era vs. $94,938 in the post-implementation era (a 94% decrease). The total cost to the Health System was $6,374,496 in the pre-implementation era vs. $2,271,003 in the postimplementation era (a 64% decrease; Table 1).
Discussion
Focusing on transparency, open communication, and ongoing quality improvement, the UMHS DA&O program, implemented in 2001, has resulted in a widespread institutional culture shift toward one of accountability and correction of systems errors that ultimately affect patient safety and the overall quality of medical care. The impact of this program on system-wide claims rates and costs has been previously analyzed (3). Our study
represents the first specialty-specific analysis of the effectiveness of the DA&O program as it pertains to GI-specific claims and costs. We detected a statistically significant decrease in the proportion of encounters resulting in claims in the post-implementation era, even in the context of a dramatic increase in clinical volume. Also notable was the marked decrease in time to claim resolution in the two eras, which has the potential to not only reduce the costs related to protracted litigation, but also hasten achievement of a sense of closure for both physicians and patients, who experience significant emotional hardship during the litigation process. This ultimately may have the effect of preserving and strengthening patient–provider relationships. Before our study, it remained an open question whether, and to what extent, the system-wide results would be reproducible in the subspecialty of GI. Neither the procedure-based nature of the field and the large number of open-access procedures performed nor our management of unique and complex patient populations such as those with functional bowel disorders appeared to influence the success of the program. In fact, our results were even more dramatic than those seen system-wide, particularly in the reduction of time to claim resolution ( > 50% in our study vs. ~30% systemwide). Specifically considering open-access procedures within GI, $700,000 $600,000
Pre-implementation Post-implementation
$500,000 $400,000
500,000
n =2
$300,000 400,000 $200,000 300,000 72% Increase
200,000
$100,000 0
n =17
n =13 n =15 Procedure-related
100,000
411,944
238,911 0 Pre-implementation Era
Post-implementation Era
n =8
n =11 Improper management/ treatment/monitoring
Delayed diagnosis/ misdiagnosis
Figure 4. Total liability dollars per claim: pre-implementation vs. post-implementation.
Figure 3. Gastroenterology (GI) patient encounters in the pre-implementation vs. post-implementation eras.
Time to claim resolution 1.00
0.75
Table 1. Total liability dollars (adjusted for inflation) in the pre-implementation vs. post-implementation eras Pre-implementation
Post-implementation
% Change
Settlement dollars
$4,743,726
$2,176,065
− 54%
Claims-related costs
$1,630,770
$94,938
− 94%
Total liability
$6,374,496
$2,271,003
− 64%
Mean total liability per claim
$167,309
$81,107
− 52%
P < 0.0001 0.50
0.25
0.00 0
500
1,000
1,500
2,000
2,500
Time (days) Before policy
The American Journal of GASTROENTEROLOGY
After policy
Figure 5. Kaplan–Meier analysis: time to claim resolution pre-implementation vs. post-implementation.
VOLUME 109 | APRIL 2014 www.amjgastro.com
THE RED SECTION
it could be inferred from our results that a pre-existing or established patient–provider relationship is not necessarily needed for this program to be successful, as long as there is transparency and open acknowledgement of medical error when it occurs. The fact that we were unable to demonstrate a significant difference between the mean total dollars incurred per claim following implementation was primarily because of the small number of total claims, resulting in an underpowered study. In addition, there is wide variability in the size of the monetary payouts, largely because of inherent differences in the specific claims, preventing standardization. One important observation was the significant decrease in nonsettlement-related expenses post-implementation, which suggests that a larger proportion of overall dollars were allocated to compensate injured patients rather than toward claimsrelated expenses. Because claims data were de-identified, we were unable to determine whether individual claims related to inpatient or outpatient care, including open-access endoscopy. It is important to recognize that our study does not definitively prove causality. Changes in practice patterns within GI during the study period, including improvements in screening/surveillance for colorectal cancer and procedural safety, may have affected the results. Furthermore, our data were institution-specific, and did not account for temporal trends in Statewide or nationwide claims during the study period. The most recent data from Michigan, compiled from professional liability insurer reporting between 2000 and 2007, show a steady decline in the number of reported claims Statewide over that period (4). However, the time from claim reporting to resolution Statewide was an average of 4.5 years, significantly longer than that reported in our study. National data reflect a slight decrease in the number of closed claims between 1999 and 2008 (2,601 vs. 2,284) (5). Adjusted for inflation, there was a 9.5% increase nationwide in the average indemnity payment for claims closed between 1999 and 2008, along with a 38% increase in average expense payments per claim. This stands in stark contrast to our post-implementation period, wherein we observed a pronounced decrease in both settlement dollars and claims-related costs. An important consideration is whether the positive results achieved following implementation of our DA&O program are generalizable to health care institutions nationwide, or even to regional medical institutions with differing physician employment structures. UMHS is self-insured, such that there exists no conflict of interest between the individual physician and the hospital system because of full indemnification. Although differences in state law governing medical malpractice may certainly influence claims rates and costs, to our knowledge the only major changes to medical malpractice legislation in Michigan during the study period were implemented in 1994. These included a compulsory 6-month pre-suit notice period requiring a plaintiff to provide written specifics of the intended claims before filing suit, which potentially served to facilitate collaborative claims investigation (6). A recent study examining stakeholders’ perceived barriers to implementation of DA&O programs categorized these barriers into four groups: cultural barriers such as physicians’ discomfort with disclosure and apology and fears about greater liability exposure, © 2014 by the American College of Gastroenterology
legal barriers such as physicians’ fears of name-based reporting of malpractice settlements and variability in governing state laws, logistical obstacles such as practice variability and complexities of inter-insurer coordination, and political barriers including ensuring favorable public policies/legislation and reassuring the public that such a system serves their best interests (7). Indeed, it is vital to recognize that a DA&O program is not, at its essence, a claims management strategy. Instead, it represents a systemic shift in medical culture toward one with a fundamental goal of improving patient safety (2). Reflecting the success of this approach, UMHS has seen an improvement in its overall safety culture scores since program implementation (2). A 1994 study surveying the reasons that patients and their surrogates take legal action found that although the opportunity for monetary compensation was certainly part of the decision-making process, other prominent driving forces included the desire for an advocate to help find answers, seek accountability, and obtain an apology, and to prevent similar errors in the future (8). For 37% of respondents, an apology and explanation would have affected their decision to file a claim. This may explain why a system aimed at open disclosure, and focused on medical error prevention rather than assigning blame results in decreased claims rates and costs, as well as expedited claims resolution. Our study confirms that DA&O programs can result in decreased medical liability claims and costs, even in a procedurebased specialty such as GI. When properly implemented as part of a system-wide quality improvement effort, such programs not only help foster a culture of transparency, and in doing so strengthen patients’ trust in the health care system, but also help to promote ongoing medical error prevention in a way that does not lead to increased medical liability. Further research should evaluate the generalizability of these findings to other practice environments. In conclusion, implementation of a DA&O program not only decreased the number of GI-related claims per patient encounter, but also dramatically shortened the time to claim resolution. This approach has the potential not only to reduce costs but also to strengthen patient–provider relationships and promote patient safety through investing in quality improvement. ACKNOWLEDGMENTS
We thank Susan Anderson for assisting with our risk management database search, Jeffrey Holden for his help in obtaining divisional data regarding clinical volume, and Richard Boothman for offering helpful insight into the evolution and application of the UMHS medical error disclosure program. CONFLICT OF INTEREST
Guarantor of the article: James M. Scheiman, MD, FACG. Specific author contributions: Megan A. Adams: study design, acquisition of data, analysis and interpretation of data, statistical analysis, drafting the manuscript; B. Joseph Elmunzer: statistical analysis and critical revision of the manuscript for important intellectual content; James M. Scheiman: study concept and design, analysis and interpretation of data, critical revision of the manuscript for important intellectual content, and study supervision. The American Journal of GASTROENTEROLOGY
463
464
THE RED SECTION
Financial support: None. Potential competing interests: Megan A. Adams and B. Joseph Elmunzer declare no conflict of interest. James M. Scheiman has served as consultant for AstraZeneca Pharmaceuticals POZEN, Stryker, McNeil Pharmaceuticals, Boehringer Ingelheim GmbH, Xlumena, and Prostrakan. REFERENCES 1. Boothman RC, Blackwell AC, Campbell DA et al. A better approach to medical malpractice claims? The University of Michigan experience. J Health Life Sci Law 2009;2:125–59. 2. Boothman RC, Imhoff SH, Campbell DA. Nurturing a culture of patient safety and achieving lower malpractice risk through disclosure: lessons learned and future directions. Front Health Serv Manage 2012;28: 13–28.
The American Journal of GASTROENTEROLOGY
3. Kachalia A, Kaufman SR, Boothman RC et al. Liability claims and costs before and after implementation of a medical error disclosure program. Ann Intern Med 2010;153:213–21. 4. Office of Financial and Insurance Regulation. Evaluation of the Michigan Medical Professional Liability Insurance Market. A Market Evaluation by Commissioner Ken Ross, Lansing, MI: State of Michigan Department of Energy, Labor, and Economic Growth; October 2009. Accessed at http:// www.michigan.gov/documents/dleg/Michigan_Medical_Liability_Ins_ Rpt_297694_7.pdf on 8 April 2013. 5. Physician Insurers Association of America. Claims Trend Analysis. A Comprehensive Analysis of Medical Liability Data Reported to the PIAA Data Sharing Project. Physician Insurers Association of America: Rockville, MD, 2009. 6. MICH. COMP LAWS § 600.2912b. 7. Bell SK, Smulowitz PB, Woodward AC et al. Disclosure, apology, and offer programs: stakeholders’ views of barriers to and strategies for broad implementation. Milbank Q 2012;90:682–705. 8. Vincent C, Young M, Phillips A. Why do people sue doctors? A study of patients and relatives taking legal action. Lancet 1994;343:1609–13.
VOLUME 109 | APRIL 2014 www.amjgastro.com
THE RED SECTION
nature publishing group
Effect of a Health System’s Medical Error Disclosure Program on Gastroenterology-Related Claims Rates and Costs Megan A. Adams, MD, JD1, B. Joseph Elmunzer, MD1 and James M. Scheiman, MD, FACG1 OBJECTIVES:
In 2001, the University of Michigan Health System (UMHS) implemented a novel medical error disclosure program. This study analyzes the effect of this program on gastroenterology (GI)-related claims and costs.
METHODS:
This was a review of claims in the UMHS Risk Management Database (1990–2010), naming a gastroenterologist. Claims were classified according to pre-determined categories. Claims data, including incident date, date of resolution, and total liability dollars, were reviewed. Mean total liability incurred per claim in the pre- and post-implementation eras was compared. Patient encounter data from the Division of Gastroenterology was also reviewed in order to benchmark claims data with changes in clinical volume.
RESULTS:
There were 238,911 GI encounters in the pre-implementation era and 411,944 in the postimplementation era. A total of 66 encounters resulted in claims: 38 in the pre-implementation era and 28 in the post-implementation era. Of the total number of claims, 15.2% alleged delay in diagnosis/misdiagnosis, 42.4% related to a procedure, and 42.4% involved improper management, treatment, or monitoring. The reduction in the proportion of encounters resulting in claims was statistically significant (P = 0.001), as was the reduction in time to claim resolution (1,000 vs. 460 days) (P < 0.0001). There was also a reduction in the mean total liability per claim ($167,309 pre vs. $81,107 post, 95% confidence interval: 33682.5–300936.2 pre vs. 1687.8–160526.7 post).
CONCLUSIONS:
Implementation of a novel medical error disclosure program, promoting transparency and quality improvement, not only decreased the number of GI-related claims per patient encounter, but also dramatically shortened the time to claim resolution.
Am J Gastroenterol 2014; 109:460–464; doi:10.1038/ajg.2013.375
Medical error disclosure, or “disclosure, apology and offer” (DA&O), programs have been implemented as novel strategies to decrease medical liability claims and costs, while also promoting disclosure, transparency, and institutional quality improvement (1). In 2001, the University of Michigan Health System (UMHS) implemented a DA&O program that revolutionized the claims management landscape of the Health System and has served as a nationwide model for liability reform. The details of this program have been reported previously (1,2). In brief, although before 2001 the UMHS employed a traditional “deny and defend” approach to medical malpractice claims, as of July 2001 the UMHS implemented a DA&O program designed to efficiently respond to incidents of clear medical error with open disclosure, an offer of compensation, and efforts at quality
improvement. At the same time, UMHS continued to vigorously defend itself against claims when it was determined that reasonable care was provided. A system-wide analysis in 2010 found a 36% decrease in the monthly rate of new claims per 100,000 patient encounters, 30% decrease in median time to claim resolution, and 44% decrease in average cost per lawsuit following program implementation (3). However, no discipline-specific studies have yet been undertaken to determine whether these findings are reproducible on an individual specialty level, as there are clearly wide variations in the scope of practice between specialties. Our study aimed to expand on the findings of the 2010 analysis by examining the effect of the UMHS DA&O program specifically on GI-related claims and costs.
1
Division of Gastroenterology, Department of Internal Medicine, University of Michigan Medical School, Ann Arbor, Michigan, USA. Correspondence: Megan A. Adams, MD, JD, Division of Gastroenterology, Department of Internal Medicine, University of Michigan, 3912 Taubman Center, SPC 5362, 1500 East Medical Center Drive, Ann Arbor, Michigan 48109, USA. E-mail: [email protected] The American Journal of GASTROENTEROLOGY
VOLUME 109 | APRIL 2014 www.amjgastro.com
THE RED SECTION
Methods
Following UMHS Institutional Review Board approval, we performed a retrospective review of de-identified claims data in the UMHS Risk Management Database between 1990 and 2010, naming a gastroenterologist. We hypothesized that, similar to the global analysis, the program would result in a significant reduction in time to claim resolution and costs with respect to GI-specific claims. We also explored whether this effect was specific to certain categories of claims. For the purposes of this study, a “claim” was defined as an assertion of wrongdoing that forms the basis for a request for compensation. Claims were independently divided into pre-determined categories: (i) delay in diagnosis/misdiagnosis, (ii) procedure-related (including sedation, nonsedation, and consent-related), and (iii) improper management, treatment, or monitoring. Any disagreement in categorization was resolved by consensus. Data regarding each claim, including time to claim resolution, disposition, and total settlement dollars and expenses, were reviewed. Mean total dollars incurred per claim were compared between the 10-year pre-implementation and 10-year post-implementation eras. These costs were subdivided into payments to patients (settlement and/or jury awards) and administrative/legal costs (nonsettlement-related expenses). Divisional patient encounter data were also reviewed in order to benchmark claims data with changes in clinical volume. As the number of inpatient encounters was relatively static (13,736 inpatient units of
Delayed diagnosis/ misdiagnosis of GI-related issue
15.2% 42.4%
GI procedural complications (sedation, nonsedation, and consent-related claims)
42.4% Improper management/ treatment/monitoring
Figure 1. Distribution of claims by category (1990–2010). GI, gastroenterology.
care in 2000 vs. 12,617 in 2009), a mean of the total number of procedures plus clinic visits was used to estimate clinical activity. Statistical methods Because expenditures in the two eras were not normally distributed (Skewness/Kurtosis test for normality, P < 0.001), the mean expenditures per claim were compared using the two-sample Wilcoxon rank-sum (Mann–Whitney) test. The difference in the proportion of encounters resulting in a claim in the two eras was analyzed using Fisher’s exact test (two-sided significance level 0.05). Time to claim resolution was compared using a Kaplan– Meier analysis.
Results
A total of 66 claims met inclusion criteria and were analyzed in the study. Of these, 15.2% alleged a delay in diagnosis/misdiagnosis, 42.4% related to a GI procedure, and another 42.4% involved improper management, treatment, or monitoring (Figure 1). In examining the distribution of claims by category in the two eras, several trends emerged. First, the percentage of claims alleging delayed diagnosis/misdiagnosis was lower in the postimplementation era than the pre-implementation era (21.1% vs. 7.1%, P = 0.007). Second, there were a larger percentage of procedure-related claims post-implementation (34.2% vs. 53.6%, P = 0.33; Figure 2). The breakdown of total procedure-related claims was as follows: 12 colonoscopy, 8 endoscopic retrograde cholangiopancreatography, 5 esophagogastroduodenoscopy, 2 capsule endoscopy, and 1 liver biopsy. There were 238,911 GI encounters in the pre-implementation era vs. 411,944 post-implementation (Figure 3). A total of 66 encounters resulted in claims, 38 pre-implementation and 28 post-implementation (despite a 72% increase in clinical activity), representing a significant reduction in the proportion of encounters resulting in claims (P = 0.001). The rate of claims per 1,000 patient encounters decreased from 0.160% pre-implementation to 0.068% post-implementation, as did the mean total liability per claim ($167,309 pre vs. $81,107 post, 95% confidence interval: 33682.5–300936.2 pre vs. 1687.8–160526.7 post; Table 1).
Pre-implementation
21.1%
Post-implementation
GI procedural complications
44.7%
34.4%
7.1%
Delayed diagnosis/ misdiagnosis
Improper management/ treatment/ monitoring
39.3% 53.6%
Figure 2. Distribution of claims by category in the pre-implementation vs. post-implementation eras. GI, gastroenterology.
© 2014 by the American College of Gastroenterology
The American Journal of GASTROENTEROLOGY
461
462
THE RED SECTION
In examining mean total liability per claim by claim type, there was a notable reduction post-implementation in both the procedure-related and improper management/treatment/monitoring categories (Figure 4). There was a marked increase in mean total liability per claim in the delayed diagnosis/misdiagnosis category due to a single outlier claim. The first quartile estimate of time to claim resolution for claims in the pre-implementation era was roughly twice that for claims filed post-implementation (1,000 vs. 460 days; P < 0.0001; Figure 5). Total settlement dollars were $4,743,726 in the pre-implementation era vs. $2,176,065 in the post-implementation era (a 54% decrease). Similarly, claims-related expenses (nonsettlement-related, including attorney’s fees and expenses from outside counsel) were $1,630,770 in the pre-implementation era vs. $94,938 in the post-implementation era (a 94% decrease). The total cost to the Health System was $6,374,496 in the pre-implementation era vs. $2,271,003 in the postimplementation era (a 64% decrease; Table 1).
Discussion
Focusing on transparency, open communication, and ongoing quality improvement, the UMHS DA&O program, implemented in 2001, has resulted in a widespread institutional culture shift toward one of accountability and correction of systems errors that ultimately affect patient safety and the overall quality of medical care. The impact of this program on system-wide claims rates and costs has been previously analyzed (3). Our study
represents the first specialty-specific analysis of the effectiveness of the DA&O program as it pertains to GI-specific claims and costs. We detected a statistically significant decrease in the proportion of encounters resulting in claims in the post-implementation era, even in the context of a dramatic increase in clinical volume. Also notable was the marked decrease in time to claim resolution in the two eras, which has the potential to not only reduce the costs related to protracted litigation, but also hasten achievement of a sense of closure for both physicians and patients, who experience significant emotional hardship during the litigation process. This ultimately may have the effect of preserving and strengthening patient–provider relationships. Before our study, it remained an open question whether, and to what extent, the system-wide results would be reproducible in the subspecialty of GI. Neither the procedure-based nature of the field and the large number of open-access procedures performed nor our management of unique and complex patient populations such as those with functional bowel disorders appeared to influence the success of the program. In fact, our results were even more dramatic than those seen system-wide, particularly in the reduction of time to claim resolution ( > 50% in our study vs. ~30% systemwide). Specifically considering open-access procedures within GI, $700,000 $600,000
Pre-implementation Post-implementation
$500,000 $400,000
500,000
n =2
$300,000 400,000 $200,000 300,000 72% Increase
200,000
$100,000 0
n =17
n =13 n =15 Procedure-related
100,000
411,944
238,911 0 Pre-implementation Era
Post-implementation Era
n =8
n =11 Improper management/ treatment/monitoring
Delayed diagnosis/ misdiagnosis
Figure 4. Total liability dollars per claim: pre-implementation vs. post-implementation.
Figure 3. Gastroenterology (GI) patient encounters in the pre-implementation vs. post-implementation eras.
Time to claim resolution 1.00
0.75
Table 1. Total liability dollars (adjusted for inflation) in the pre-implementation vs. post-implementation eras Pre-implementation
Post-implementation
% Change
Settlement dollars
$4,743,726
$2,176,065
− 54%
Claims-related costs
$1,630,770
$94,938
− 94%
Total liability
$6,374,496
$2,271,003
− 64%
Mean total liability per claim
$167,309
$81,107
− 52%
P < 0.0001 0.50
0.25
0.00 0
500
1,000
1,500
2,000
2,500
Time (days) Before policy
The American Journal of GASTROENTEROLOGY
After policy
Figure 5. Kaplan–Meier analysis: time to claim resolution pre-implementation vs. post-implementation.
VOLUME 109 | APRIL 2014 www.amjgastro.com
THE RED SECTION
it could be inferred from our results that a pre-existing or established patient–provider relationship is not necessarily needed for this program to be successful, as long as there is transparency and open acknowledgement of medical error when it occurs. The fact that we were unable to demonstrate a significant difference between the mean total dollars incurred per claim following implementation was primarily because of the small number of total claims, resulting in an underpowered study. In addition, there is wide variability in the size of the monetary payouts, largely because of inherent differences in the specific claims, preventing standardization. One important observation was the significant decrease in nonsettlement-related expenses post-implementation, which suggests that a larger proportion of overall dollars were allocated to compensate injured patients rather than toward claimsrelated expenses. Because claims data were de-identified, we were unable to determine whether individual claims related to inpatient or outpatient care, including open-access endoscopy. It is important to recognize that our study does not definitively prove causality. Changes in practice patterns within GI during the study period, including improvements in screening/surveillance for colorectal cancer and procedural safety, may have affected the results. Furthermore, our data were institution-specific, and did not account for temporal trends in Statewide or nationwide claims during the study period. The most recent data from Michigan, compiled from professional liability insurer reporting between 2000 and 2007, show a steady decline in the number of reported claims Statewide over that period (4). However, the time from claim reporting to resolution Statewide was an average of 4.5 years, significantly longer than that reported in our study. National data reflect a slight decrease in the number of closed claims between 1999 and 2008 (2,601 vs. 2,284) (5). Adjusted for inflation, there was a 9.5% increase nationwide in the average indemnity payment for claims closed between 1999 and 2008, along with a 38% increase in average expense payments per claim. This stands in stark contrast to our post-implementation period, wherein we observed a pronounced decrease in both settlement dollars and claims-related costs. An important consideration is whether the positive results achieved following implementation of our DA&O program are generalizable to health care institutions nationwide, or even to regional medical institutions with differing physician employment structures. UMHS is self-insured, such that there exists no conflict of interest between the individual physician and the hospital system because of full indemnification. Although differences in state law governing medical malpractice may certainly influence claims rates and costs, to our knowledge the only major changes to medical malpractice legislation in Michigan during the study period were implemented in 1994. These included a compulsory 6-month pre-suit notice period requiring a plaintiff to provide written specifics of the intended claims before filing suit, which potentially served to facilitate collaborative claims investigation (6). A recent study examining stakeholders’ perceived barriers to implementation of DA&O programs categorized these barriers into four groups: cultural barriers such as physicians’ discomfort with disclosure and apology and fears about greater liability exposure, © 2014 by the American College of Gastroenterology
legal barriers such as physicians’ fears of name-based reporting of malpractice settlements and variability in governing state laws, logistical obstacles such as practice variability and complexities of inter-insurer coordination, and political barriers including ensuring favorable public policies/legislation and reassuring the public that such a system serves their best interests (7). Indeed, it is vital to recognize that a DA&O program is not, at its essence, a claims management strategy. Instead, it represents a systemic shift in medical culture toward one with a fundamental goal of improving patient safety (2). Reflecting the success of this approach, UMHS has seen an improvement in its overall safety culture scores since program implementation (2). A 1994 study surveying the reasons that patients and their surrogates take legal action found that although the opportunity for monetary compensation was certainly part of the decision-making process, other prominent driving forces included the desire for an advocate to help find answers, seek accountability, and obtain an apology, and to prevent similar errors in the future (8). For 37% of respondents, an apology and explanation would have affected their decision to file a claim. This may explain why a system aimed at open disclosure, and focused on medical error prevention rather than assigning blame results in decreased claims rates and costs, as well as expedited claims resolution. Our study confirms that DA&O programs can result in decreased medical liability claims and costs, even in a procedurebased specialty such as GI. When properly implemented as part of a system-wide quality improvement effort, such programs not only help foster a culture of transparency, and in doing so strengthen patients’ trust in the health care system, but also help to promote ongoing medical error prevention in a way that does not lead to increased medical liability. Further research should evaluate the generalizability of these findings to other practice environments. In conclusion, implementation of a DA&O program not only decreased the number of GI-related claims per patient encounter, but also dramatically shortened the time to claim resolution. This approach has the potential not only to reduce costs but also to strengthen patient–provider relationships and promote patient safety through investing in quality improvement. ACKNOWLEDGMENTS
We thank Susan Anderson for assisting with our risk management database search, Jeffrey Holden for his help in obtaining divisional data regarding clinical volume, and Richard Boothman for offering helpful insight into the evolution and application of the UMHS medical error disclosure program. CONFLICT OF INTEREST
Guarantor of the article: James M. Scheiman, MD, FACG. Specific author contributions: Megan A. Adams: study design, acquisition of data, analysis and interpretation of data, statistical analysis, drafting the manuscript; B. Joseph Elmunzer: statistical analysis and critical revision of the manuscript for important intellectual content; James M. Scheiman: study concept and design, analysis and interpretation of data, critical revision of the manuscript for important intellectual content, and study supervision. The American Journal of GASTROENTEROLOGY
463
464
THE RED SECTION
Financial support: None. Potential competing interests: Megan A. Adams and B. Joseph Elmunzer declare no conflict of interest. James M. Scheiman has served as consultant for AstraZeneca Pharmaceuticals POZEN, Stryker, McNeil Pharmaceuticals, Boehringer Ingelheim GmbH, Xlumena, and Prostrakan. REFERENCES 1. Boothman RC, Blackwell AC, Campbell DA et al. A better approach to medical malpractice claims? The University of Michigan experience. J Health Life Sci Law 2009;2:125–59. 2. Boothman RC, Imhoff SH, Campbell DA. Nurturing a culture of patient safety and achieving lower malpractice risk through disclosure: lessons learned and future directions. Front Health Serv Manage 2012;28: 13–28.
The American Journal of GASTROENTEROLOGY
3. Kachalia A, Kaufman SR, Boothman RC et al. Liability claims and costs before and after implementation of a medical error disclosure program. Ann Intern Med 2010;153:213–21. 4. Office of Financial and Insurance Regulation. Evaluation of the Michigan Medical Professional Liability Insurance Market. A Market Evaluation by Commissioner Ken Ross, Lansing, MI: State of Michigan Department of Energy, Labor, and Economic Growth; October 2009. Accessed at http:// www.michigan.gov/documents/dleg/Michigan_Medical_Liability_Ins_ Rpt_297694_7.pdf on 8 April 2013. 5. Physician Insurers Association of America. Claims Trend Analysis. A Comprehensive Analysis of Medical Liability Data Reported to the PIAA Data Sharing Project. Physician Insurers Association of America: Rockville, MD, 2009. 6. MICH. COMP LAWS § 600.2912b. 7. Bell SK, Smulowitz PB, Woodward AC et al. Disclosure, apology, and offer programs: stakeholders’ views of barriers to and strategies for broad implementation. Milbank Q 2012;90:682–705. 8. Vincent C, Young M, Phillips A. Why do people sue doctors? A study of patients and relatives taking legal action. Lancet 1994;343:1609–13.
VOLUME 109 | APRIL 2014 www.amjgastro.com
