Implementation of the Management of Anticoagulation in the Periprocedural Period App into an Electronic Health Record: A Cost-Effectiveness Analysis

Abstract

The Management of Anticoagulation in the Periprocedural Period (MAPPP) app is a free tool providing up-to-date guidelines on the periprocedural management of patients on long-term anticoagulants. After validating its effectiveness in the post-procedural period, we aimed to study its overall cost-effectiveness. SF-12 surveys were sent to eligible patients, converted into SF-6D forms, and subsequently into quality-adjusted life years (QALYs) to calculate the incremental cost-effectiveness ratio (ICER). The number of 30-day readmissions was used to calculate hospitalization costs, utilizing publicly available data. From 1/1/2018 to 1/31/2019, 642 patients were screened for enrollment, with an overall response rate of 94% (164/175) among the consented and 49% (164/336) among all eligible patients. The average QALY score was 0.7134 (95% CI [0.6836, 0.7431]) for the patients whose treatment plan followed the MAPPP app recommendations (acceptance group) and 0.7104 (95% CI [0.6760, 0.7448]) for those who did not (rejection group), without statistically significant differences. The difference in ICER scores was −$429 866.67, with the negative sign demonstrating that acceptance was the dominant strategy. By utilizing QALYs and ICER scores we have shown that the acceptance of MAPPP app recommendations is the dominant strategy for the periprocedural management of patients on long-term anticoagulation.

Keywords

MAPPP periprocedural management anticoagulation cost-effectiveness analysis electronic health record health informatics

Introduction

The recent advent and accessibility of online and mobile technology have resulted in a plethora of e-health products created by the medical community, many of which aim to assist physicians in clinical decision-making, while empowering transparency and patient involvement.¹ Many physicians rely heavily on accessing up-to-date medical information by using mobile applications on-the-go during their practice. A common use case of this paradigm is chronic conditions with rapidly changing treatment guidelines, such as cardiovascular diseases, where the use of mobile apps and wearable devices has been increasingly gaining interest from both academic centers and industrial stakeholders worldwide.² This is even more important when considering the potential cost-effectiveness of such implementations, given the low cost of owning a modern mobile device and the easily accessible online health-related resources. In a recent systematic review of the cost-effectiveness of digital health interventions for the management of cardiovascular diseases (including short message services, telephone support, mobile applications, video conferencing systems, digital transmission of physiologic data, and wearable medical devices), the use of digital health interventions was deemed cost-effective.³

The periprocedural management of patients on long-term anticoagulation is a unique case study of these strategies, where timely and up-to-date evidence-based guidance is crucial in assisting practicing physicians in their efforts to balance the risk of bleeding from anticoagulation against the risk of thrombosis from its discontinuation. The Management of Anticoagulation in the Periprocedural Period (MAPPP) mobile app is a free clinical decision support tool developed by Island Peer Review Organization, Inc. (IPRO), the Centers for Medicare & Medicaid Services designated Quality Innovation Network–Quality Improvement Organization for the state of New York, which provides the most current guidance of the periprocedural management of patients on long-term oral anticoagulation.⁴ MAPPP has previously been studied for its uptake and utilization by health care providers⁵ and was proven to be effective in the reduction of 30-day readmissions and deaths.⁶ The positive outcomes of the MAPPP app gave ground for a further analysis of its effects on cost reduction as well as cost-effectiveness in this group of patients.

We, therefore, decided to perform analyses for the MAPPP app tool's effects on cost reduction as well as cost-effectiveness in order to evaluate its potential for further expansion within a greater user base in the US healthcare system and beyond.

Methods

Patients and Data Acquisition

Patient demographics and clinical data were obtained from our institution's electronic health record (EHR) system (Allscripts Healthcare Solutions, Inc., Chicago, IL). An SF-12 survey was sent to eligible patients who had previously participated in our prospective cohort regarding the effectiveness of the MAPPP app.⁶ The study was approved by the health system Institutional Review Board (IRB#:18-0948) overseeing all hospitals and healthcare facilities visited by the patients and was conducted in accordance with principles of the Declaration of Helsinki. Informed consent forms were given to all patients who were invited to complete the survey.

Survey Process

Email invitations were sent to eligible patients containing a link to the online survey form and included an informed consent paragraph (Figure 1). For patients who did not respond, email reminders were sent biweekly, for a total of three reminders per patient. Patients who did not respond after these reminders were contacted by our staff through phone calls in order to perform the survey over the phone, with or without proxies, or schedule a more convenient time to do so. At any time, the clinical staff were blinded and kept a log of the calls made to each patient and, as a last resort, a paper form was sent to their address. Strenuous effort was made to document any reason for refusal or inability to participate.

Figure 1.

Flowchart of the survey process.

The quality-adjusted life years (QALYs) are the academic standard for measuring how well different kinds of medical treatments lengthen and/or improve patients’ lives.⁷ The answered SF-12 survey forms were first converted into the SF-6D form, which allowed their subsequent conversion into QALYs based on an available mathematical formula.^8–10 The participants were separated into two groups depending on whether their healthcare providers had previously decided to accept the app's treatment recommendations (acceptance group) or not (rejection group), based on the patient cohorts previously defined.⁶ The total number of 30-day hospital readmissions of the responders was used to calculate the related costs for each group. These costs and the resulting quality-of-life (QOL) data were used to calculate the incremental cost-effectiveness ratio (ICER).¹¹ Healthcare-related cost data were retrieved from the Centers for Medicare & Medicaid Services (CMS) Medicare Beneficiaries at a Glance report (2021 edition), a dashboard published by the US government which is publicly available online.¹²

Statistical Analyses

Descriptive frequency and percentage data were calculated for categorical variables to show patient characteristics between the surveyed group and the non-surveyed group, as well as between the acceptance group and the rejection group among surveyed patients. For continuous variables (age, weight, creatinine clearance, and serum creatinine), the mean and standard deviation (SD) were calculated for each group. Chi-square tests (or Fisher's Exact test for having a cell with less than five counts) were used for categorical variable comparison and student-t tests were used to compare continuous variables, including QALYs. P-values < .05 were considered statistically significant. SAS v9.4 (SAS, Cary, NC) was used for all statistical analyses.

Results

During the study period of 1/1/2018 to 1/31/2019, there were 642 eligible patients for enrollment (Figure 2). Of these, consent could not be obtained from 306 (163 were unable to be reached; 73 had a missing procedure date; 46 lacked contact information; 15 expired; 9 did not speak English). A total of 336 patients remained that qualified for consenting. Of those 336 patients, 175 consented while 161 were unwilling to take the survey. Additionally, out of the 175 that consented, 11 did not complete the survey, leading to an overall consent rate of 52% (175/336). The overall response rate was 94% (164/175) among the consented and 49% (164/336) among eligible patients.

Figure 2.

Flow diagram of patient enrollment.

Of the total 642 patients, 164 completed the survey and 478 did not. There were no statistical differences for any demographic (age, sex, weight) or clinical (creatinine clearance, use of anticoagulation medications, use of antiplatelet medications) factors between the surveyed group and the non-surveyed group (Table 1), confirming our assumption that there was no selection bias in the survey sample. Of the 164 patients in the survey group, the mean age was 71.1; 45% were female. Mean weight, creatinine clearance, and serum creatinine were 89.8 kg, 83.1 mL/min, and 1.2 mg/dL, respectively. Among this group, 40% were on warfarin, 32.9% were on apixaban, 24.4% were on rivaroxaban, and 2.4% were on dabigatran. Regarding the use of antiplatelet medications, 15.9% were on aspirin, and 3.1% on clopidogrel. Considering the surgical operations that the patients had undergone, 53.7% had a high procedure bleeding risk, 40.2% had a low bleed risk procedure, and 6.1% had a minimal bleed risk procedure.

Table 1.

Patient Characteristics at Baseline by Survey Completion status (N = 642).

Patient characteristics	Surveyed (n = 164)	Not surveyed (n = 478)	P-value
Age, mean (SD)	71.11 (13.26)	73.12 (12.32)	.0772
Sex			.4353
Male	90 (54.88%)	279 (58.37%)
Female	74 (45.12%)	199 (41.63%)
Weight kg, mean (SD)	89.76 (26.30)	85.59 (23.16)	.0783
Creatinine clearance ml/min, mean (SD)	83.06 (53.34)	74.00 (38.03)	.0512
Creatinine mg/dl, mean (SD)	1.16 (0.58)	1.25 (1.42)	.2467
Anticoagulation medication			.1881
Warfarin	66 (40.24%)	166 (34.73%)
Dabigatran	4 (2.44%)	31 (6.49%)
Rivaroxaban	40 (24.39%)	116 (24.27%)
Apixaban	54 (32.93%)	165 (34.52%)
Antiplatelet medication
Aspirin			.7928
Yes	26 (15.85%)	80 (16.74%)
No	138 (84.15%)	398 (83.26%)
Clopidogrel			.5167
Yes	5 (3.05%)	20 (4.18%)
No	159 (96.95%)	458 (95.82%)
Procedure bleeding risk			.8678
Minimal	10 (6.10%)	31 (6.49%)
Low	66 (40.24%)	202 (42.26%)
High	88 (53.66%)	245 (51.26%)
Patient's thromboembolic risk			.089
Low	80 (48.78%)	196 (41%)
Medium	65 (39.63%)	196 (41%)
High	19 (11.59%)	86 (17.99%)
Intervention group			.1546
Acceptance	98 (59.76%)	255 (53.35%)
Rejection	66 (40.24%)	223 (46.65%)

Note: Due to the presence of missing values, the actual final number of patients with values for weight, creatinine clearance and serum creatinine was slightly different than the number of patients listed in the table.

Of the 164 patients who took the SF-12 survey, 98 were in the acceptance group while 66 were in the rejection group. When the responses of the two groups were compared to each other by each question, no statistically significant differences were found (Table 2), indicating that none of the survey questions were associated with the MAPPP app intervention acceptance status.

Table 2.

Survey Results by Tool Acceptance status (N = 164).

Survey questions	Accepted (n = 98)	Rejected (n = 66)	P-value
In general, would you say your health is			.2334
Excellent	6 (6.12)	5 (7.58)
Very Good	23 (23.47)	9 (13.64)
Good	37 (37.76)	36 (54.55)
Fair	21 (21.43)	11 (16.67)
Poor	11 (11.22)	5 (7.58)
Did you notice any limitation of moderate activities such as moving a table, pushing a vacuum cleaner, bowling, or playing golf?			.6102
YES, limited a lot	22 (33.33)	27 (27.55)
YES, limited a little	21 (31.82)	38 (38.78)
NO, not limited at all	23 (34.85)	33 (33.67)
Did you notice any limitation in climbing several flights of stairs?			.8635
YES, limited a lot	35 (35.71)	21 (31.82)
YES, limited a little	36 (36.73)	25 (37.88)
NO, not limited at all	27 (27.55)	20 (30.3)
Did you accomplish less than you would like (physically)?			.6847
Yes	61 (62.24)	39 (59.9)
No	37 (37.76)	27 (40.91)
Did you feel limited in the kind of work or other activities?			.5217
Yes	59 (60.2)	43 (65.15)
No	39 (39.8)	23 (34.85)
Did you accomplish less than you would like (emotionally)?			.4515
Yes	29 (29.59)	16 (24.24)
No	69 (70.41)	50 (75.76)
Did you perform work or other activities less carefully than usual?			.0517
Yes	20 (20.41)	6 (9.09)
No	78 (79.59)	60 (90.91)
During the past 4 weeks, how much did pain interfere with your normal work (including work outside the home and housework)?			.3857
Not at all	35 (35.71)	22 (33.33)
A little bit	20 (20.41)	18 (27.27)
Moderately	22 (22.45)	18 (27.27)
Quite a bit	15 (15.31)	4 (6.06)
Extremely	6 (6.12)	4 (6.06)
Have you felt calm & peaceful?			.4778
All of the time	12 (12.24)	15 (22.73)
Most of the time	43 (43.88)	28 (42.42)
A good bit of the time	10 (10.20)	7 (10.61)
Some of the time	23 (23.47)	13 (19.70)
A little of the time	7 (7.14)	2 (3.03)
None of the time	3 (3.06)	1 (1.52)
Did you have a lot of energy?			.847
All of the time	6 (6.12)	2 (3.03)
Most of the time	24 (24.49)	16 (24.24)
A good bit of the time	11 (11.22)	7 (10.61)
Some of the time	38 (38.78)	24 (36.36)
A little of the time	11 (11.22)	8 (12.12)
None of the time	8 (8.16)	9 (13.64)
Have you felt downhearted and blue?			.1027
All of the time	1 (1.02)	2 (3.03)
Most of the time	9 (9.18)	1 (1.52)
A good bit of the time	3 (3.06)	0 (0.00)
Some of the time	25 (25.51)	12 (18.18)
A little of the time	28 (28.57)	23 (34.85)
None of the time	32 (32.65)	28 (42.42)
During the past 4 weeks, how much of the time has your physical health or emotional problems interfered with your social activities (like visiting friends, relatives, etc)?			.699
All of the time	4 (4.08)	6 (9.09)
Most of the time	12 (12.24)	7 (10.61)
A good bit of the time	0 (0.00)	0 (0.00)
Some of the time	27 (27.55)	15 (22.73)
A little of the time	17 (17.35)	13 (19.70)
None of the time	38 (38.78)	25 (37.88)

The average QALY score was 0.7134 (95% CI [0.6836, 0.7431]) for the acceptance group and 0.7104 (95% CI [0.6760, 0.7448]) for the rejection group, showing no statistically significant difference between these two groups in terms of QALYs (Table 3), despite the higher QALY score of the acceptance group compared to the rejection group. Based on the data from CMS Medicare Beneficiaries at a Glance the average cost for a hospital stay is $14 234. Since there were three readmissions in the acceptance group and 8 in the rejection group, the average readmission cost for each group was $435.70 and $1725.30, respectively. Assuming that other related costs are negligible in each group compared to the cost of hospitalization, the ICER score was calculated as (435.7–1725.3)/(0.7134–0.7104) = −429 866.67, the negative sign of which demonstrates that acceptance was the dominant strategy.

Table 3.

QALY and Costs by Tool Acceptance status (N = 164).

Item	Accepted (n = 98)	Rejected (n = 66)	P-value
QALY Mean [95% CI]	0.7134 [0.6836, 0.7431]	0.7104 [0.6760, 0.7448]	.8777
Average cost of hospital stays in CMS	$14 234	$14 234	.0288^a
Total readmissions	3	8
Total Cost	$42 702	$113 872
Average Cost in each group	$435.70	$1725.30
ICER score	($435.70-$1725.30)/(0.7134–0.7104) = −$429 866.67

Abbreviations: CI, Confidence Intervals; CMS, Centers for Medicare & Medicaid Services; ICER, Incremental Cost-Effectiveness Ratio.

Fisher's exact test is used here.

Discussion

When a new intervention is both clinically superior and cost saving, it is referred to as an economically “dominant” strategy, the opposite is a “dominated” strategy.¹³ Our study revealed that by utilizing the QALY and ICER scoring systems we have shown that the acceptance of MAPPP app recommendations is the dominant, cost-effective strategy for the periprocedural management of patients on long-term anticoagulation compared to not following the app's recommendations. Approximately 15%–20% of patients on long-term oral anticoagulant therapy will require a surgical procedure each year,¹⁴ which highlights the clinical significance of the MAPPP app and the need for broader adoption of the tool. The difficult decision of selecting the correct plan for periprocedural antithrombotic management commonly falls to the hands of medical residents or pre-surgical staff, who understandably may not have received dedicated training targeted towards this complex patient group. The recently released perioperative antithrombotic management guidelines¹⁴ included the MAPPP app as an interactive and easy to use tool, potentially bridging this knowledge gap by providing the necessary guidance when needed. Our study supports funding for the MAPPP app by stakeholders to accelerate its integration into an official clinical workflow in EHR systems and to provider point-of-care systems.

Cost-effectiveness analyses have been used before in the field of anticoagulation for the purpose of guiding policies by using similar tools such as QALYs and ICER.^13,15 If evidence shows that a treatment helps lengthen life or improve QOL, these benefits are comprehensively summed up to calculate how many additional QALYs the treatment provides, and this added health benefit is then compared to the added health benefit of other treatments for the same patient population.⁷ In our study, the overall consent rate for the MAPPP survey was 52%, with a response rate of 94% among the consented and 49% among all eligible patients. Given that part of the study was conducted during the early phases of the COVID-19 pandemic the survey had to be completed remotely, with no direct patient contact. The fact that most patients were over 65 years old meant that many of them faced difficulties with completing the online surveys by themselves. We believe that these factors contributed to the number of patients that were unwilling to participate, which is moderately common in similar studies with older participants that are conducted remotely. In terms of our cost-effectiveness analysis, the average QALY score for the acceptance group was minimally higher than for the rejection group, while the average cost for the acceptance group was slightly lower, leading to an overall negative ICER. This final calculation confirmed that the acceptance of the tool's recommendations is the dominant strategy in terms of cost-effectiveness.

The relatively small number of patients in our study might have contributed to the non-statistically significant difference of QALYs among the surveyed patients. Larger ecological studies have shown that hospitalizations caused by venous thromboembolism led to a significant loss of disability-adjusted life years (DALYs) worldwide.¹⁶ Our analysis demonstrates that the use of MAPPP has a clear financial benefit affecting both the beneficiaries and the governmental sponsors, as these funds can be allocated to improving other aspects of healthcare, especially after the massive resource utilization in the healthcare sector caused by the COVID-19 pandemic.

Our study had several limitations. Aside from the readmission cost, we did not have in our possession a detailed list of other related costs (such as software installation and intervention personnel cost) for each group, albeit they are minimal compared to the costs of hospitalization, since hospitalization is usually the biggest component of the total healthcare-associated costs. Although the information offered from the CMS website lacks detailed descriptions of the individual costs, it is a publicly available database validated by the federal government. This makes it a reliable source of information that can be easily accessed by policy makers to guide future decisions based on the results of our study. We also chose not to calculate long-term (1-year) estimates of related costs, based on the fact that the majority of thrombotic and other adverse outcomes in patients undergoing surgeries usually occur within four weeks from the day of the procedure, with subsequent risk gradually decreasing after that time period.^17,18 Since the primary diagnoses of most readmissions that occurred were related to venous thromboembolism, it is expected that the most significant effect on both patient QALYs and the related costs would be seen in the early post-discharge phase. A longer follow-up period would offer little in our analysis and at the same time would have increased the technical difficulties of the study, considering that it was completed during the early phases of the COVID-19 pandemic.

Conclusion

After a cost-effectiveness analysis using a patient survey and publicly available data from governmental sources, our results show that the acceptance of the MAPPP app's recommendations is the dominant strategy for the periprocedural management of patients on long-term anticoagulation. By utilizing the QALY and ICER scoring systems, our study has validated the MAPPP app's utility and cost-effectiveness during the initial evaluation phase, indicating that it is mature enough to be used in a daily clinical EHR workflow.

Footnotes

Acknowledgements

This was produced collaboratively with IPRO, the Centers for Medicare & Medicaid Services (CMS) designated Quality Improvement Network Quality Improvement Organization (QIN-QIO) for New York State, and lead for the Atlantic Quality Innovation Network (AQIN) under the 11th Statements of Work. The analyses on which this publication is based were performed under Contract Number HHSM-500-2014-QIN013I, funded by CMS, an agency of the US Department of Health and Human Services (HHS). The content of this publication does not necessarily reflect the views or policies of the HHS nor does mention of trade names, commercial products, or organizations imply endorsement by the US government. The authors assume full responsibility for the accuracy and completeness of the ideas presented.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by A.C.S. reports honoraria from Janssen, Bayer, Bristol Myers Squibb, Pfizer, and Sanofi and research grants from Janssen and Boehringer Ingelheim. He is also a member of the ATLAS group, an academic research organization.

ORCID iDs

Jason J. Wang

Ioannis Koulas

Alex C. Spyropoulos

References

Watson

Tribe

Shennan

. The role of medical smartphone apps in clinical decision-support: a literature review. Artif Intell Med. 2019;100:101707. doi:10.1016/j.artmed.2019.101707.

Ventura

Sousa

Dixe

, et al. A clinical decision support system for remote monitoring of cardiovascular disease patients: a clinical study protocol. Front Public Health. 2022;10:859890. doi:10.3389/fpubh.2022.859890.

Jiang

Ming

You

. The cost-effectiveness of digital health interventions on the management of cardiovascular diseases: systematic review. J Med Internet Res. 2019;21(6):e13166. doi:10.2196/13166.

IPRO. Management of Anticoagulation in the Peri-Procedural Period (MAPPP) app. https://mappp.ipro.org/. Accessed August 31, 2022.

Spyropoulos

Myrka

Triller

, et al. Uptake and utilization of the management of anticoagulation in the periprocedural period app: longitudinal analysis. JMIR Mhealth Uhealth. 2018;6(12):e11090. doi:10.2196/11090.

Spyropoulos

Giannis

Cohen

, et al. Implementation of the management of anticoagulation in the periprocedural period app into an electronic health record: a prospective cohort study. Clin Appl Thromb Hemost. 2020;26:1076029620925910. doi:10.1177/1076029620925910.

ICER. Cost-Effectiveness, the QALY, and the evLYG. https://icer.org/our-approach/methods-process/cost-effectiveness-the-qaly-and-the-evlyg/.

Brazier

Roberts

Deverill

. The estimation of a preference-based measure of health from the SF-36. J Health Econ. 2002;21(2):271–292. doi:10.1016/S0167-6296(01)00130-8.

Kharroubi

Brazier

Roberts

O'Hagan

. Modelling SF-6D health state preference data using a nonparametric Bayesian method. J Health Econ. 2007;26(3):597–612. doi:10.1016/j.jhealeco.2006.09.002.

10.

Brazier

Roberts

. The estimation of a preference-based measure of health from the SF-12. Med Care. 2004;42(9):851–859.

11.

Sanders

Neumann

Basu

, et al. Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: second panel on cost-effectiveness in health and medicine. JAMA. 2016;316(10):1093–1103. doi:10.1001/jama.2016.12195.

12.

CMS Medicare Beneficiaries at a Glance, 2021 EDITION. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Beneficiary-Snapshot/Bene_Snapshot. Accessed May 6, 2022.

13.

Cohen

Reynolds

. Interpreting the results of cost-effectiveness studies. J Am Coll Cardiol. 2008;52(25):2119–2126. doi:10.1016/j.jacc.2008.09.018.

14.

Douketis

Spyropoulos

Murad

, et al. Perioperative management of antithrombotic therapy: an American college of chest physicians clinical practice guideline. Chest. 2022. doi:10.1016/j.chest.2022.07.025.

15.

Carlson

Kuderer

, et al. Cost-effectiveness analysis of low-dose direct oral anticoagulant (DOAC) for the prevention of cancer-associated thrombosis in the United States. Cancer. 2020;126(8):1736–1748. doi:10.1002/cncr.32724.

16.

ISTH Steering Committee for World Thrombosis Day. Thrombosis: a major contributor to global disease burden. Thromb Res. 2014;134(5):931–938. doi:10.1016/j.thromres.2014.08.014.

17.

Sweetland

Green

Liu

, et al. Duration and magnitude of the postoperative risk of venous thromboembolism in middle aged women: prospective cohort study. Br Med J. 2009;339:b4583. doi:10.1136/bmj.b4583.

18.

Amin

Deitelzweig

Bucior

, et al. Frequency of hospital readmissions for venous thromboembolism and associated hospital costs and length of stay among acute medically ill patients in the US. J Med Econ. 2019;22(11):1119–1125. doi:10.1080/13696998.2019.1618862.