ORIGINAL ARTICLE Myocardial perfusion imaging determination using an appropriate use smartphone application Ashish Mahajan, MD,a Susan Bal, MD,a and Harvey Hahn, MDa a
Department of Internal Medicine, Kettering Medical Center, Kettering, OH
Received May 31, 2014; accepted Sep 3, 2014 doi:10.1007/s12350-014-9995-0
Background. Inappropriate cardiac imaging has been a significant cost concern and cause of radiation burden to patients. Objective. To assess if a smartphone application (app) based on 2009 Appropriate Use Criteria (AUC) for Cardiac Radionuclide Imaging published by American College of Cardiology would be feasible at the point of order. Methods. We evaluated stress myocardial perfusion imaging (MPI) (N 5 403) (mean age 5 62.23 years; 47.89% males) over a 4 month period using a smartphone app to determine whether the study ordered was Appropriate, Inappropriate, or Uncertain per 2009 AUC. We also monitored the time needed to use the app to determine the level of appropriateness of each stress MPI. The results of the stress MPI were noted. Results. Of the 403 stress MPIs evaluated, 267 (66.25%) were noted to be Appropriate, 118 (29.28%) were Inappropriate, and 13 (3.23%) were Uncertain, per AUC; 5 (1.25%) remained unclassified. Average time needed to use the app to assess each stress MPI for appropriateness was noted to be 44 (±9) seconds. Non-teaching physicians ordered 70 (38.89%) inappropriate stress MPIs as compared to 20 (23.53%) ordered by physicians on resident teaching service, and 28 (23.33%) by cardiologists (P 5 .0045). Among inappropriately ordered stress MPIs, 87 (42.65%) were ordered in females as compared to 31 (17.13%) in males (P < .0001). 70 (26.22%) stress MPIs among appropriately ordered were abnormal (reversible ischemia or fixed perfusion defect) as compared to 15 (12.17%) among inappropriately ordered stress MPIs (P 5 .0032). Conclusion. A free and convenient smartphone app provides an easy-to-use tool to assist physicians in determining the level of appropriateness of stress MPI in a time- and cost-effective manner at the point of order. The smartphone app may have potential to promote the usage of the AUC and possibly aid reduction of healthcare cost and ionizing radiation burden. (J Nucl Cardiol 2015;22:66–71.) Key Words: Appropriate Æ smartphone Æ app Æ stress Æ radiation
INTRODUCTION With the advent of smartphones and tablets, there has been increasing trend among physicians to incorporate these devices into their daily practice. Smartphone
Reprint requests: Ashish Mahajan, MD, Department of Internal Medicine, Kettering Medical Center, 3535 Southern Boulevard, Kettering, OH 45429; [email protected] 1071-3581/$34.00 Copyright Ó 2014 American Society of Nuclear Cardiology. 66
enables the physician to access clinical applications, evidence based resources, and advanced mobile communication in one handheld-sized device at the point of care. Smartphones applications (commonly known as apps) have become ubiquitous in many aspects of our lives over past few years and healthcare is rapidly adopting this technology. Recognizing increasing popularity of smartphones and apps in healthcare, multiple web-based tools and smartphone apps have been developed to assist physicians in improving patient care.
Journal of Nuclear CardiologyÒ Volume 22, Number 1;66–71
Mahajan et al Myocardial perfusion imaging determination
67
The smartphone app based on 2009 AUC is free of cost and can be easily downloaded on all major smartphone operating systems. The app is essentially a smartphone version of the AUC document. It provides all 67 indications of AUC in a stepwise decision tree format. The initial screen provides the user, options to view AUC in either an algorithm or indication category format. The app allows the user to access the AUC tables in a hierarchical fashion. It then navigates the user to select an indication category based on pertinent clinical information. The app also provides a built-in Framingham risk score (FRS) calculator as well as a table to determine pretest probability of coronary artery disease. The app does not need any patient information to be physically entered into it thus making it HIPPA compliant, however requires the user to know pertinent clinical information about the patient.
staff. This hospital is located in a suburb of the Dayton Metropolitan area in Ohio. All stress MPIs included in the study were performed at this hospital. No prior authorization was required for stress MPI. The study was approved by Institutional Review Board of the institution. Inpatients between 18 and 89 years age undergoing stress MPI over a 4-month period (Nov 2013 to Jan 2014) were included in the study. Outpatients were excluded due to inability to access pertinent clinical information. The pertinent clinical information of each patient included in the study was obtained using electronic medical record on the same day after the stress MPI was performed. Demographic factors, presenting symptoms, risk factors, prior medical history including prior coronary intervention were obtained. The ordering physicians were classified into three categories, physicians on teaching service, non-teaching physicians, and cardiologists. None of the ordering physicians were aware that the stress MPI studies were being scrutinized for appropriateness and they were not utilizing the app prior to ordering a stress MPI. There were no major differences in patients seen by teaching and non-teaching physicians, except for patients requiring pre-operative medical clearance, which were seen by non-teaching physicians. For patients admitted by Internal Medicine Service, where cardiology consultation was obtained, the physician physically ordering the stress MPI was assigned. The studies ordered by residents, fellows, physician assistants, and nurse practitioners were classified based on the specialty of the supervising physician. The smartphone app was then used to determine whether the study ordered was Appropriate, Inappropriate, or Uncertain. Patients who could be classified under more than one category, the consensus was reached based on independent evaluation by third investigator as well as discussion among the investigators. The FRS calculator built into the app was used to calculate FRS for risk assessment. The time needed to use the app to determine the level of appropriateness and select the indication category was recorded by two investigators. One of the investigators used a stopwatch to record the time taken, while the second investigator used the app. The time recorded was from the moment the investigator clicked into the app to determine appropriateness and does not include the time needed to review patient’s chart and collect pertinent patient information. The results of the stress MPI were noted. A stress MPI was classified as abnormal if there was either reversible ischemia or fixed perfusion defect. The inappropriately ordered studies which were abnormal were followed up. ECAM Siemens Dual Head camera was used for each stress MPI. Tc-99m Sestamibi was used with 30 mCi for rest images and 10 mCi for stress images. Filtered back projection processing was applied and Butterworth filter was used for filtering. No attenuation correction was applied. The software program used to determine ejection fraction was 4DMSPECT. No scoring system is being used at our institute to interpret the studies. Only single observer interpreted an individual stress MPI. All interpreters were blinded to the AUC ratings.
Study Design
Statistical Analysis
The study was performed at a teaching community hospital with residents, fellows, teaching and non-teaching physician
The variables used in statistical analysis are listed in Table 1.
Stress myocardial perfusion imaging (MPI) has remained the cornerstone of non-invasive diagnostic cardiac imaging. Inappropriate stress MPI is a major healthcare cost concern over number of years. There also has been a growing concern regarding radiation burden from stress MPI studies.1 The American College of Cardiology (ACC) developed Appropriate Use Criteria (AUC) for cardiac radionuclide imaging in 2005, which were subsequently updated in 20092 and most recently as multimodality AUC in 2013.3 Multiple studies have evaluated the applicability of AUC in various clinical settings. Ever since the publication of AUC, the ACC has been making efforts to promote their usage through quality improvement initiatives such as FOCUS campaign4 and Choosing Wisely initiative5 with an aim to ultimately reduce inappropriate testing. The AUC although comprehensive, are fairly voluminous, making it difficult to be used at the point of order. A well-known pharmaceutical company developed a smartphone app based on 2009 AUC.6 This app is an easily accessible tool comprising all AUC indications in smartphone format. The app, therefore, brings the entire AUC to the physicians’ pocket for use at the point of order. To authors’ knowledge, no previous study has evaluated the utility of a smartphone app in determining appropriateness and its potential role in promoting usage of AUC. The aim of our study was to utilize this smartphone app to evaluate stress MPI studies performed at our institution for appropriateness and also to record the time needed to use the app. MATERIALS AND METHODS Appropriate Use Criteria Smartphone App
68
Mahajan et al Myocardial perfusion imaging determination
Journal of Nuclear CardiologyÒ January/February 2015
Table 1. Study variables
Variable
N
Sex Female 210 Male 193 Physician group Cardiologist 127 Non-teaching 187 Teaching 89 AUC Appropriate 267 Inappropriate 118 Uncertain 13 Unclassified 5 Stress MPI result Negative 313 Positive 90
Table 2. Baseline characteristics
Percentage
95% CI
52.11 47.89
(47.21, 57.01) (42.99, 52.79)
31.51 46.40 22.08
(26.96, 36.07) (41.51, 51.29) (18.02, 26.15)
66.25 29.28 3.23 1.25
(61.62, 70.89) (24.82, 33.74) (1.49, 4.96) (0.16, 2.33)
77.67 22.33
(73.58, 81.75) (18.25, 26.42)
N Mean age (years) Sex Females Males BMI [ 30 Diabetes mellitus (Type 1 or 2) Hypertension Hyperlipidemia Coronary artery disease Prior PCI/CABG Family history of CAD Current smoker
403 62.23 (SD = 14.01) 52.11% 47.89% 49.62% 30.52% 71.46% 66.50% 26.80% 17.86% 47.39% 27.29%
N, Total study population; CAD, Coronary artery disease; PCI, Percutaneous coronary intervention; CABG, Coronary artery bypass grafting.
CI, Confidence interval; AUC, Appropriate use criteria.
Table 3. Classification of study based on stress MPI indication
The data used to answer the research questions included only those studies that were classified as appropriate or inappropriate. Pearson v2 test was used for the analysis. P value \ .05 was considered significant. All analyses were conducted using SAS version 9.3 (Cary, NC).
Chest pain Non-angina Atypical Typical Dyspnea Pre-operative Syncope EKG changes Palpitations Positive troponins
RESULTS Baseline Characteristics The baseline characteristics of the study population are shown in Table 2. Classification of patients based on stress MPI indication is shown in Table 3.
population
64.51% 51.53% 35.38% 13.07% 18.85% 7.19% 3.97% 1.98% 1.73% 1.74%
Level of Appropriateness 267 (66.25%) studies were noted to be appropriate, 118 (29.28%) were inappropriate, and 13 (3.23%) studies were uncertain per AUC. 5 (1.25%) studies remained unclassified. These results are illustrated in Figure 1. Indications Based on Appropriateness Of the 267 appropriate stress MPI studies, 121 (45.31%) were ordered for symptomatic patients presenting with acute chest pain with possible acute coronary syndrome with no EKG changes and negative troponins, 64 (23.97%) were ordered for evaluation of symptomatic patients presenting with non-acute Ischemic Equivalent, 35 (13.10%) for evaluation of new
Figure 1. Appropriateness Category.
symptoms in patients with history of prior revascularization (PCI or CABG). The rest of the appropriate stress MPI studies were ordered for other appropriate indications per 2009 AUC document. The symptoms
Journal of Nuclear CardiologyÒ Volume 22, Number 1;66–71
Mahajan et al Myocardial perfusion imaging determination
69
classified as Ischemic Equivalent are as per defined in the AUC document. Of the 118 inappropriate stress MPI studies, 61 (51.69%) were ordered for evaluation of non-acute chest pain with low pre-test probability of CAD with interpretable EKG and ability to exercise, 24 (20.33%) for risk assessment in patients undergoing non-cardiac surgery, 14 (11.86%) for patients with stable symptoms but prior history of coronary artery disease or abnormal prior stress imaging study, but no revascularization.
Stress MPI Physician Ordering Patterns The differences in appropriateness of stress MPI among different physician groups is illustrated in Figure 2. We classified physicians ordering the stress MPI into physicians on resident teaching service, nonteaching physicians, and cardiologists. We noted that non-teaching physicians ordered 70 (38.89%) inappropriate stress MPIs as compared to 20 (23.53%) ordered by physicians on resident teaching service and 28 (23.33%) by cardiologists. The difference between nonteaching physicians as compared to physicians on resident teaching service and cardiologists was statistically significant (P = .0045).
Figure 2. Appropriateness differences among physician groups.
Gender Differences We noted that among inappropriately ordered stress MPI studies, 87 (42.65%) were ordered in females as compared to 31 (17.13%) in males. The difference was statistically significant (P \ .0001).
Time Needed to Determine Appropriateness Average time needed to assess each stress MPI for appropriateness using the app by investigators was noted to be 44(±9) seconds.
Appropriateness and Stress MPI Results The results of stress MPI and its relation to appropriateness are illustrated in Figure 3. Of the total 403 stress MPI studies, only 85 (21.09%) were found to be abnormal. Of the 85 abnormal studies, 70 (26.22%) were appropriately ordered whereas 15 (12.17%) were inappropriately ordered. The difference was statistically significant (P = .0032). Of the 70 appropriately ordered studies, which were abnormal, 37 (52.85%) had reversible ischemia and 33 (47.14%) had fixed perfusion defect.
Figure 3. Relation of appropriateness to stress MPI results.
Of the 15 inappropriately ordered studies, which were abnormal, 8 (53.33%) had reversible ischemia, whereas 7 (46.66%) had fixed perfusion defect. Follow Up Inappropriately Ordered Abnormal MPI Studies Of the 15 abnormal inappropriately ordered stress MPI studies, 11 patients underwent cardiac catheterization. Of these 11 patients, 6 (54.54%) had significant CAD and required revascularization, whereas 5 (45.45%) did not have significant CAD. DISCUSSION The aim of our study was to assess the utility of a smartphone app to determine appropriateness of stress
70
Mahajan et al Myocardial perfusion imaging determination
MPI at the point of order. We found that an AUC based smartphone app can be conveniently used by physicians at the point of order to determine the level of appropriateness. The use of the app required less than a minute to determine appropriateness, thereby demonstrating decision support in a time effective manner. To our knowledge, this is the first study to assess the utility of a smartphone app based on AUC to determine appropriateness. In this era of technology, smartphones and tablets are rapidly gaining popularity in healthcare and most physicians are looking to find ways to integrate these into their daily practice beyond checking their emails. The rate of smartphone usage among physicians has been reported as high as 81% by Jackson and Cocker research associates,7 75% by Payne et al.8 and Franko et al. reported that 56% of physicians including medical students, residents, and attending physicians were using medical apps in their practice.9 In our healthcare reform-minded environment, apps that enable fast and easy access to AUC are becoming increasingly important. A recent study done by Lin et al. demonstrated rapid determination of test appropriateness for CAD evaluation using an AUC based point-of-order decision support tool.10 Given the increasing popularity of smartphone apps, our study has a potential role in promoting usage of AUC among healthcare professionals. This may potentially reduce inappropriate stress testing and further translate into huge healthcare cost saving as well as reduced radiation exposure from unnecessary testing. We noted a higher rate of inappropriateness (29.28%) as compared to most prior studies. Previously, the rate of inappropriateness of stress MPI has been reported as low as 7% to as high as 24.2%. Recently, Doukky et al. reported highest percentage of inappropriate stress MPIs at 45.5%.11 Most recently, Singh et al. reported lowest percentage of inappropriate stress MPI studies at 5.5%.12 Notably, most prior studies based on 2009 AUC were done either at tertiary care facilities located in Asia or University hospitals within the United States with exception of study performed by Singh et al. in a rural tertiary care setting and Doukky et al. in an office-based setting. In contrast, being performed in a community hospital setting in the US, our study may reflect overutilization pattern of stress MPI in community hospitals. We found a higher proportion of inappropriately ordered stress MPI studies by non-teaching physicians (38.89%) as compared to physicians on resident teaching service (23.53%) and cardiologists (23.33%). This is in contrast to Singh et al.12 and Carryer et al.,13 who reported no significant differences in terms of appropriateness between cardiologists and non-cardiologists. The
Journal of Nuclear CardiologyÒ January/February 2015
higher proportion of inappropriately ordered studies by non-teaching physicians as compared to physicians on resident teaching service and cardiologists may be attributed to the fact that most of the stress MPI studies ordered for pre-operative testing were inappropriate and these patients were primarily seen by non-teaching physicians. However, this may also demonstrate awareness and superior adherence to AUC guidelines among cardiologists and physicians on resident teaching service due to their teaching responsibilities. The higher proportion of inappropriately ordered stress MPIs in females (42.65%) as compared to males (17.13%) noted in our study was consistent with prior studies. These gender differences in inappropriate stress testing may be due to the fact that for similar symptoms and risk factors, females may have low pretest probability of CAD as compared to males, resulting in higher proportion of stress MPIs classified as inappropriate in females. The higher proportion of positive stress MPIs among appropriately ordered (26.22%) as compared to inappropriately ordered stress MPIs (12.71%) noted in our study was in concordance with Khwaja et al., who reported 40% positive among appropriate vs 18% among inappropriate.14 The significant proportion of abnormal results among inappropriately ordered stress MPI studies suggests that test abnormality may not correspond to appropriateness. LIMITATIONS First, the study was performed using an app based on 2009 AUC, which uses the old terminologies such as ‘‘Appropriate’’, ‘‘Inappropriate’’, and ‘‘Uncertain’’ that may have a limited use after the recent introduction of new terminologies ‘‘Appropriate’’, ‘‘May be Appropriate’’, and ‘‘Rarely Appropriate’’. Second, the smartphone app was used by our investigators, who had some expertise in operating the app before the study. The reproducibility of the time effectiveness of the app when operated on large scale by other physicians needs further evaluation. Third, this was a single-center study and the higher rates of inappropriateness reported in our study as compared to most previous studies may not reflect nationwide trends. Fourth, we did not assess utility of other less expensive non-invasive tests such as stress echocardiogram or exercise treadmill test in patients with inappropriately ordered stress MPI. Lastly, we did not include outpatients undergoing stress MPI in our study due to lack of accessibility of pertinent clinical information, which may potentially skew the appropriateness of the results.
Journal of Nuclear CardiologyÒ Volume 22, Number 1;66–71
Mahajan et al Myocardial perfusion imaging determination
NEW KNOWLEDGE GAINED Smartphone app can be conveniently used at the point of order in a time effective manner to promote the usage of AUC among physicians. This could have an impact on medical costs as well as radiation burden. CONCLUSION A free and convenient smartphone app provides an easy-to-use tool to assist physicians to determine the level of appropriateness of stress MPI at the point of order in less than a minute. The smartphone app may have potential to promote the usage of AUC and possibly aid reduction of healthcare cost and ionizing radiation burden. Disclosures No grants, contracts, and other forms of financial support were provided from any source.
4. 5. 6. 7. 8.
9. 10.
References 1. Einstein AndrewJ. Effects of radiation exposure from cardiac imaging: How good are the data? J Am Coll Cardiol 2012;59:553– 65. 2. Hendel RC, Berman DS, Di Carli MF, Heidenreich PA, Henkin RE, Pellikka PA, et al. ACCF/ASNC/ACR/AHA/ASE/SCCT/ SCMR/SNM 2009 appropriate use criteria for cardiac radionuclide imaging: A report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, the American Society of Nuclear Cardiology, the American College of Radiology, the American Heart Association, the American Society of Echocardiography, the Society of Cardiovascular Computed Tomography, the Society for Cardiovascular Magnetic Resonance, and the Society of Nuclear Medicine. Circulation 2009;119:561–87. 3. Ronan G, Wolk MJ, Bailey SR, Doherty JU, Douglas PS, Hendel RC, et al. ACCF/AHA/ASE/ASNC/HFSA/HRS/SCAI/SCCT/ SCMR/STS 2013 multimodality appropriate use criteria for the
11.
12.
13.
14.
71
detection and risk assessment of stable ischemic heart disease: A report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, and Society of Thoracic Surgeons. J Nucl Cardiol 2014;21:192–220. http://www.cardiosource.org/science-and-quality/quality-programs/ imaging-in-focus.aspx. Accessed 4 Jan 2014. http://www.choosingwisely.org/wp-content/uploads/2012/12/CW_ 5Things_ACC.pdf. Accessed 16 Dec 2013. http://www.astellasapps.com/Webapp.aspx. Accessed 1 Sept 2013. http://industryreport.jacksoncoker.com//physician-career-resources/ newsletters/monthlymain/des/apps.aspx. Accessed 3 Dec 2013. Payne KB, Wharrad H, Watts K. Smartphone and medical related App use among medical students and junior doctors in the United Kingdom (UK): A regional survey. BMC Med Inf Decis Mak 2012;12:121. Franko OI, Tirrell TF. Smartphone app use among medical providers in ACGME training programs. J Med Syst 2012;36:3135–9. Lin FY, Dunning AM, Narula J, Shaw LJ, Gransar H, Berman DS, et al. Impact of an automated multimodality point-of-order decision support tool on rates of appropriate testing and clinical decision making for individuals with suspected coronary artery disease: A prospective multicenter study. J Am Coll Cardiol 2013;62:308–16. Doukky R, Hayes K, Frogge N, Balakrishnan G, Dontaraju VS, Rangel MO, et al. Impact of appropriate use on the prognostic value of single-photon emission computed tomography myocardial perfusion imaging. Circulation 2013;128:1634–43. Singh M, Babayan Z, Harjai KJ, Dedhia P, Sattur S, Jagasia DH, et al. Utilization patterns of single-photon emission cardiac tomography myocardial perfusion imaging studies in a rural tertiary care setting. Clin Cardiol 2014;37:67–72. Carryer DJ, Askew JW, Hodge D, Miller TD, Gibbons RJ. The impact of ordering provider specialty on appropriateness classification. J Nucl Cardiol 2012;19:285–90. Khawaja FJ, Jouni H, Miller TD, Hodge DO, Gibbons RJ. Downstream clinical implications of abnormal myocardial perfusion single-photon emission computed tomography based on appropriate use criteria. J Nucl Cardiol. 2013;20:1041–8.
Department of Internal Medicine, Kettering Medical Center, Kettering, OH
Received May 31, 2014; accepted Sep 3, 2014 doi:10.1007/s12350-014-9995-0
Background. Inappropriate cardiac imaging has been a significant cost concern and cause of radiation burden to patients. Objective. To assess if a smartphone application (app) based on 2009 Appropriate Use Criteria (AUC) for Cardiac Radionuclide Imaging published by American College of Cardiology would be feasible at the point of order. Methods. We evaluated stress myocardial perfusion imaging (MPI) (N 5 403) (mean age 5 62.23 years; 47.89% males) over a 4 month period using a smartphone app to determine whether the study ordered was Appropriate, Inappropriate, or Uncertain per 2009 AUC. We also monitored the time needed to use the app to determine the level of appropriateness of each stress MPI. The results of the stress MPI were noted. Results. Of the 403 stress MPIs evaluated, 267 (66.25%) were noted to be Appropriate, 118 (29.28%) were Inappropriate, and 13 (3.23%) were Uncertain, per AUC; 5 (1.25%) remained unclassified. Average time needed to use the app to assess each stress MPI for appropriateness was noted to be 44 (±9) seconds. Non-teaching physicians ordered 70 (38.89%) inappropriate stress MPIs as compared to 20 (23.53%) ordered by physicians on resident teaching service, and 28 (23.33%) by cardiologists (P 5 .0045). Among inappropriately ordered stress MPIs, 87 (42.65%) were ordered in females as compared to 31 (17.13%) in males (P < .0001). 70 (26.22%) stress MPIs among appropriately ordered were abnormal (reversible ischemia or fixed perfusion defect) as compared to 15 (12.17%) among inappropriately ordered stress MPIs (P 5 .0032). Conclusion. A free and convenient smartphone app provides an easy-to-use tool to assist physicians in determining the level of appropriateness of stress MPI in a time- and cost-effective manner at the point of order. The smartphone app may have potential to promote the usage of the AUC and possibly aid reduction of healthcare cost and ionizing radiation burden. (J Nucl Cardiol 2015;22:66–71.) Key Words: Appropriate Æ smartphone Æ app Æ stress Æ radiation
INTRODUCTION With the advent of smartphones and tablets, there has been increasing trend among physicians to incorporate these devices into their daily practice. Smartphone
Reprint requests: Ashish Mahajan, MD, Department of Internal Medicine, Kettering Medical Center, 3535 Southern Boulevard, Kettering, OH 45429; [email protected] 1071-3581/$34.00 Copyright Ó 2014 American Society of Nuclear Cardiology. 66
enables the physician to access clinical applications, evidence based resources, and advanced mobile communication in one handheld-sized device at the point of care. Smartphones applications (commonly known as apps) have become ubiquitous in many aspects of our lives over past few years and healthcare is rapidly adopting this technology. Recognizing increasing popularity of smartphones and apps in healthcare, multiple web-based tools and smartphone apps have been developed to assist physicians in improving patient care.
Journal of Nuclear CardiologyÒ Volume 22, Number 1;66–71
Mahajan et al Myocardial perfusion imaging determination
67
The smartphone app based on 2009 AUC is free of cost and can be easily downloaded on all major smartphone operating systems. The app is essentially a smartphone version of the AUC document. It provides all 67 indications of AUC in a stepwise decision tree format. The initial screen provides the user, options to view AUC in either an algorithm or indication category format. The app allows the user to access the AUC tables in a hierarchical fashion. It then navigates the user to select an indication category based on pertinent clinical information. The app also provides a built-in Framingham risk score (FRS) calculator as well as a table to determine pretest probability of coronary artery disease. The app does not need any patient information to be physically entered into it thus making it HIPPA compliant, however requires the user to know pertinent clinical information about the patient.
staff. This hospital is located in a suburb of the Dayton Metropolitan area in Ohio. All stress MPIs included in the study were performed at this hospital. No prior authorization was required for stress MPI. The study was approved by Institutional Review Board of the institution. Inpatients between 18 and 89 years age undergoing stress MPI over a 4-month period (Nov 2013 to Jan 2014) were included in the study. Outpatients were excluded due to inability to access pertinent clinical information. The pertinent clinical information of each patient included in the study was obtained using electronic medical record on the same day after the stress MPI was performed. Demographic factors, presenting symptoms, risk factors, prior medical history including prior coronary intervention were obtained. The ordering physicians were classified into three categories, physicians on teaching service, non-teaching physicians, and cardiologists. None of the ordering physicians were aware that the stress MPI studies were being scrutinized for appropriateness and they were not utilizing the app prior to ordering a stress MPI. There were no major differences in patients seen by teaching and non-teaching physicians, except for patients requiring pre-operative medical clearance, which were seen by non-teaching physicians. For patients admitted by Internal Medicine Service, where cardiology consultation was obtained, the physician physically ordering the stress MPI was assigned. The studies ordered by residents, fellows, physician assistants, and nurse practitioners were classified based on the specialty of the supervising physician. The smartphone app was then used to determine whether the study ordered was Appropriate, Inappropriate, or Uncertain. Patients who could be classified under more than one category, the consensus was reached based on independent evaluation by third investigator as well as discussion among the investigators. The FRS calculator built into the app was used to calculate FRS for risk assessment. The time needed to use the app to determine the level of appropriateness and select the indication category was recorded by two investigators. One of the investigators used a stopwatch to record the time taken, while the second investigator used the app. The time recorded was from the moment the investigator clicked into the app to determine appropriateness and does not include the time needed to review patient’s chart and collect pertinent patient information. The results of the stress MPI were noted. A stress MPI was classified as abnormal if there was either reversible ischemia or fixed perfusion defect. The inappropriately ordered studies which were abnormal were followed up. ECAM Siemens Dual Head camera was used for each stress MPI. Tc-99m Sestamibi was used with 30 mCi for rest images and 10 mCi for stress images. Filtered back projection processing was applied and Butterworth filter was used for filtering. No attenuation correction was applied. The software program used to determine ejection fraction was 4DMSPECT. No scoring system is being used at our institute to interpret the studies. Only single observer interpreted an individual stress MPI. All interpreters were blinded to the AUC ratings.
Study Design
Statistical Analysis
The study was performed at a teaching community hospital with residents, fellows, teaching and non-teaching physician
The variables used in statistical analysis are listed in Table 1.
Stress myocardial perfusion imaging (MPI) has remained the cornerstone of non-invasive diagnostic cardiac imaging. Inappropriate stress MPI is a major healthcare cost concern over number of years. There also has been a growing concern regarding radiation burden from stress MPI studies.1 The American College of Cardiology (ACC) developed Appropriate Use Criteria (AUC) for cardiac radionuclide imaging in 2005, which were subsequently updated in 20092 and most recently as multimodality AUC in 2013.3 Multiple studies have evaluated the applicability of AUC in various clinical settings. Ever since the publication of AUC, the ACC has been making efforts to promote their usage through quality improvement initiatives such as FOCUS campaign4 and Choosing Wisely initiative5 with an aim to ultimately reduce inappropriate testing. The AUC although comprehensive, are fairly voluminous, making it difficult to be used at the point of order. A well-known pharmaceutical company developed a smartphone app based on 2009 AUC.6 This app is an easily accessible tool comprising all AUC indications in smartphone format. The app, therefore, brings the entire AUC to the physicians’ pocket for use at the point of order. To authors’ knowledge, no previous study has evaluated the utility of a smartphone app in determining appropriateness and its potential role in promoting usage of AUC. The aim of our study was to utilize this smartphone app to evaluate stress MPI studies performed at our institution for appropriateness and also to record the time needed to use the app. MATERIALS AND METHODS Appropriate Use Criteria Smartphone App
68
Mahajan et al Myocardial perfusion imaging determination
Journal of Nuclear CardiologyÒ January/February 2015
Table 1. Study variables
Variable
N
Sex Female 210 Male 193 Physician group Cardiologist 127 Non-teaching 187 Teaching 89 AUC Appropriate 267 Inappropriate 118 Uncertain 13 Unclassified 5 Stress MPI result Negative 313 Positive 90
Table 2. Baseline characteristics
Percentage
95% CI
52.11 47.89
(47.21, 57.01) (42.99, 52.79)
31.51 46.40 22.08
(26.96, 36.07) (41.51, 51.29) (18.02, 26.15)
66.25 29.28 3.23 1.25
(61.62, 70.89) (24.82, 33.74) (1.49, 4.96) (0.16, 2.33)
77.67 22.33
(73.58, 81.75) (18.25, 26.42)
N Mean age (years) Sex Females Males BMI [ 30 Diabetes mellitus (Type 1 or 2) Hypertension Hyperlipidemia Coronary artery disease Prior PCI/CABG Family history of CAD Current smoker
403 62.23 (SD = 14.01) 52.11% 47.89% 49.62% 30.52% 71.46% 66.50% 26.80% 17.86% 47.39% 27.29%
N, Total study population; CAD, Coronary artery disease; PCI, Percutaneous coronary intervention; CABG, Coronary artery bypass grafting.
CI, Confidence interval; AUC, Appropriate use criteria.
Table 3. Classification of study based on stress MPI indication
The data used to answer the research questions included only those studies that were classified as appropriate or inappropriate. Pearson v2 test was used for the analysis. P value \ .05 was considered significant. All analyses were conducted using SAS version 9.3 (Cary, NC).
Chest pain Non-angina Atypical Typical Dyspnea Pre-operative Syncope EKG changes Palpitations Positive troponins
RESULTS Baseline Characteristics The baseline characteristics of the study population are shown in Table 2. Classification of patients based on stress MPI indication is shown in Table 3.
population
64.51% 51.53% 35.38% 13.07% 18.85% 7.19% 3.97% 1.98% 1.73% 1.74%
Level of Appropriateness 267 (66.25%) studies were noted to be appropriate, 118 (29.28%) were inappropriate, and 13 (3.23%) studies were uncertain per AUC. 5 (1.25%) studies remained unclassified. These results are illustrated in Figure 1. Indications Based on Appropriateness Of the 267 appropriate stress MPI studies, 121 (45.31%) were ordered for symptomatic patients presenting with acute chest pain with possible acute coronary syndrome with no EKG changes and negative troponins, 64 (23.97%) were ordered for evaluation of symptomatic patients presenting with non-acute Ischemic Equivalent, 35 (13.10%) for evaluation of new
Figure 1. Appropriateness Category.
symptoms in patients with history of prior revascularization (PCI or CABG). The rest of the appropriate stress MPI studies were ordered for other appropriate indications per 2009 AUC document. The symptoms
Journal of Nuclear CardiologyÒ Volume 22, Number 1;66–71
Mahajan et al Myocardial perfusion imaging determination
69
classified as Ischemic Equivalent are as per defined in the AUC document. Of the 118 inappropriate stress MPI studies, 61 (51.69%) were ordered for evaluation of non-acute chest pain with low pre-test probability of CAD with interpretable EKG and ability to exercise, 24 (20.33%) for risk assessment in patients undergoing non-cardiac surgery, 14 (11.86%) for patients with stable symptoms but prior history of coronary artery disease or abnormal prior stress imaging study, but no revascularization.
Stress MPI Physician Ordering Patterns The differences in appropriateness of stress MPI among different physician groups is illustrated in Figure 2. We classified physicians ordering the stress MPI into physicians on resident teaching service, nonteaching physicians, and cardiologists. We noted that non-teaching physicians ordered 70 (38.89%) inappropriate stress MPIs as compared to 20 (23.53%) ordered by physicians on resident teaching service and 28 (23.33%) by cardiologists. The difference between nonteaching physicians as compared to physicians on resident teaching service and cardiologists was statistically significant (P = .0045).
Figure 2. Appropriateness differences among physician groups.
Gender Differences We noted that among inappropriately ordered stress MPI studies, 87 (42.65%) were ordered in females as compared to 31 (17.13%) in males. The difference was statistically significant (P \ .0001).
Time Needed to Determine Appropriateness Average time needed to assess each stress MPI for appropriateness using the app by investigators was noted to be 44(±9) seconds.
Appropriateness and Stress MPI Results The results of stress MPI and its relation to appropriateness are illustrated in Figure 3. Of the total 403 stress MPI studies, only 85 (21.09%) were found to be abnormal. Of the 85 abnormal studies, 70 (26.22%) were appropriately ordered whereas 15 (12.17%) were inappropriately ordered. The difference was statistically significant (P = .0032). Of the 70 appropriately ordered studies, which were abnormal, 37 (52.85%) had reversible ischemia and 33 (47.14%) had fixed perfusion defect.
Figure 3. Relation of appropriateness to stress MPI results.
Of the 15 inappropriately ordered studies, which were abnormal, 8 (53.33%) had reversible ischemia, whereas 7 (46.66%) had fixed perfusion defect. Follow Up Inappropriately Ordered Abnormal MPI Studies Of the 15 abnormal inappropriately ordered stress MPI studies, 11 patients underwent cardiac catheterization. Of these 11 patients, 6 (54.54%) had significant CAD and required revascularization, whereas 5 (45.45%) did not have significant CAD. DISCUSSION The aim of our study was to assess the utility of a smartphone app to determine appropriateness of stress
70
Mahajan et al Myocardial perfusion imaging determination
MPI at the point of order. We found that an AUC based smartphone app can be conveniently used by physicians at the point of order to determine the level of appropriateness. The use of the app required less than a minute to determine appropriateness, thereby demonstrating decision support in a time effective manner. To our knowledge, this is the first study to assess the utility of a smartphone app based on AUC to determine appropriateness. In this era of technology, smartphones and tablets are rapidly gaining popularity in healthcare and most physicians are looking to find ways to integrate these into their daily practice beyond checking their emails. The rate of smartphone usage among physicians has been reported as high as 81% by Jackson and Cocker research associates,7 75% by Payne et al.8 and Franko et al. reported that 56% of physicians including medical students, residents, and attending physicians were using medical apps in their practice.9 In our healthcare reform-minded environment, apps that enable fast and easy access to AUC are becoming increasingly important. A recent study done by Lin et al. demonstrated rapid determination of test appropriateness for CAD evaluation using an AUC based point-of-order decision support tool.10 Given the increasing popularity of smartphone apps, our study has a potential role in promoting usage of AUC among healthcare professionals. This may potentially reduce inappropriate stress testing and further translate into huge healthcare cost saving as well as reduced radiation exposure from unnecessary testing. We noted a higher rate of inappropriateness (29.28%) as compared to most prior studies. Previously, the rate of inappropriateness of stress MPI has been reported as low as 7% to as high as 24.2%. Recently, Doukky et al. reported highest percentage of inappropriate stress MPIs at 45.5%.11 Most recently, Singh et al. reported lowest percentage of inappropriate stress MPI studies at 5.5%.12 Notably, most prior studies based on 2009 AUC were done either at tertiary care facilities located in Asia or University hospitals within the United States with exception of study performed by Singh et al. in a rural tertiary care setting and Doukky et al. in an office-based setting. In contrast, being performed in a community hospital setting in the US, our study may reflect overutilization pattern of stress MPI in community hospitals. We found a higher proportion of inappropriately ordered stress MPI studies by non-teaching physicians (38.89%) as compared to physicians on resident teaching service (23.53%) and cardiologists (23.33%). This is in contrast to Singh et al.12 and Carryer et al.,13 who reported no significant differences in terms of appropriateness between cardiologists and non-cardiologists. The
Journal of Nuclear CardiologyÒ January/February 2015
higher proportion of inappropriately ordered studies by non-teaching physicians as compared to physicians on resident teaching service and cardiologists may be attributed to the fact that most of the stress MPI studies ordered for pre-operative testing were inappropriate and these patients were primarily seen by non-teaching physicians. However, this may also demonstrate awareness and superior adherence to AUC guidelines among cardiologists and physicians on resident teaching service due to their teaching responsibilities. The higher proportion of inappropriately ordered stress MPIs in females (42.65%) as compared to males (17.13%) noted in our study was consistent with prior studies. These gender differences in inappropriate stress testing may be due to the fact that for similar symptoms and risk factors, females may have low pretest probability of CAD as compared to males, resulting in higher proportion of stress MPIs classified as inappropriate in females. The higher proportion of positive stress MPIs among appropriately ordered (26.22%) as compared to inappropriately ordered stress MPIs (12.71%) noted in our study was in concordance with Khwaja et al., who reported 40% positive among appropriate vs 18% among inappropriate.14 The significant proportion of abnormal results among inappropriately ordered stress MPI studies suggests that test abnormality may not correspond to appropriateness. LIMITATIONS First, the study was performed using an app based on 2009 AUC, which uses the old terminologies such as ‘‘Appropriate’’, ‘‘Inappropriate’’, and ‘‘Uncertain’’ that may have a limited use after the recent introduction of new terminologies ‘‘Appropriate’’, ‘‘May be Appropriate’’, and ‘‘Rarely Appropriate’’. Second, the smartphone app was used by our investigators, who had some expertise in operating the app before the study. The reproducibility of the time effectiveness of the app when operated on large scale by other physicians needs further evaluation. Third, this was a single-center study and the higher rates of inappropriateness reported in our study as compared to most previous studies may not reflect nationwide trends. Fourth, we did not assess utility of other less expensive non-invasive tests such as stress echocardiogram or exercise treadmill test in patients with inappropriately ordered stress MPI. Lastly, we did not include outpatients undergoing stress MPI in our study due to lack of accessibility of pertinent clinical information, which may potentially skew the appropriateness of the results.
Journal of Nuclear CardiologyÒ Volume 22, Number 1;66–71
Mahajan et al Myocardial perfusion imaging determination
NEW KNOWLEDGE GAINED Smartphone app can be conveniently used at the point of order in a time effective manner to promote the usage of AUC among physicians. This could have an impact on medical costs as well as radiation burden. CONCLUSION A free and convenient smartphone app provides an easy-to-use tool to assist physicians to determine the level of appropriateness of stress MPI at the point of order in less than a minute. The smartphone app may have potential to promote the usage of AUC and possibly aid reduction of healthcare cost and ionizing radiation burden. Disclosures No grants, contracts, and other forms of financial support were provided from any source.
4. 5. 6. 7. 8.
9. 10.
References 1. Einstein AndrewJ. Effects of radiation exposure from cardiac imaging: How good are the data? J Am Coll Cardiol 2012;59:553– 65. 2. Hendel RC, Berman DS, Di Carli MF, Heidenreich PA, Henkin RE, Pellikka PA, et al. ACCF/ASNC/ACR/AHA/ASE/SCCT/ SCMR/SNM 2009 appropriate use criteria for cardiac radionuclide imaging: A report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, the American Society of Nuclear Cardiology, the American College of Radiology, the American Heart Association, the American Society of Echocardiography, the Society of Cardiovascular Computed Tomography, the Society for Cardiovascular Magnetic Resonance, and the Society of Nuclear Medicine. Circulation 2009;119:561–87. 3. Ronan G, Wolk MJ, Bailey SR, Doherty JU, Douglas PS, Hendel RC, et al. ACCF/AHA/ASE/ASNC/HFSA/HRS/SCAI/SCCT/ SCMR/STS 2013 multimodality appropriate use criteria for the
11.
12.
13.
14.
71
detection and risk assessment of stable ischemic heart disease: A report of the American College of Cardiology Foundation Appropriate Use Criteria Task Force, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Heart Failure Society of America, Heart Rhythm Society, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, and Society of Thoracic Surgeons. J Nucl Cardiol 2014;21:192–220. http://www.cardiosource.org/science-and-quality/quality-programs/ imaging-in-focus.aspx. Accessed 4 Jan 2014. http://www.choosingwisely.org/wp-content/uploads/2012/12/CW_ 5Things_ACC.pdf. Accessed 16 Dec 2013. http://www.astellasapps.com/Webapp.aspx. Accessed 1 Sept 2013. http://industryreport.jacksoncoker.com//physician-career-resources/ newsletters/monthlymain/des/apps.aspx. Accessed 3 Dec 2013. Payne KB, Wharrad H, Watts K. Smartphone and medical related App use among medical students and junior doctors in the United Kingdom (UK): A regional survey. BMC Med Inf Decis Mak 2012;12:121. Franko OI, Tirrell TF. Smartphone app use among medical providers in ACGME training programs. J Med Syst 2012;36:3135–9. Lin FY, Dunning AM, Narula J, Shaw LJ, Gransar H, Berman DS, et al. Impact of an automated multimodality point-of-order decision support tool on rates of appropriate testing and clinical decision making for individuals with suspected coronary artery disease: A prospective multicenter study. J Am Coll Cardiol 2013;62:308–16. Doukky R, Hayes K, Frogge N, Balakrishnan G, Dontaraju VS, Rangel MO, et al. Impact of appropriate use on the prognostic value of single-photon emission computed tomography myocardial perfusion imaging. Circulation 2013;128:1634–43. Singh M, Babayan Z, Harjai KJ, Dedhia P, Sattur S, Jagasia DH, et al. Utilization patterns of single-photon emission cardiac tomography myocardial perfusion imaging studies in a rural tertiary care setting. Clin Cardiol 2014;37:67–72. Carryer DJ, Askew JW, Hodge D, Miller TD, Gibbons RJ. The impact of ordering provider specialty on appropriateness classification. J Nucl Cardiol 2012;19:285–90. Khawaja FJ, Jouni H, Miller TD, Hodge DO, Gibbons RJ. Downstream clinical implications of abnormal myocardial perfusion single-photon emission computed tomography based on appropriate use criteria. J Nucl Cardiol. 2013;20:1041–8.
