Education and electronic medical records and genomics network, challenges, and lessons learned from a large-scale clinical trial using polygenic risk scores

Polygenic risk scores (PRS) have potential to improve health care by identifying individuals that have elevated risk for common complex conditions. Use of PRS in clinical practice, however, requires careful assessment of the needs and capabilities of patients, providers, and health care systems. The electronic Medical Records and Genomics (eMERGE) network is conducting a collaborative study which will return PRS to 25,000 pediatric and adult participants. All participants will receive a risk report, potentially classifying them as high risk (∼2-10% per condition) for 1 or more of 10 conditions based on PRS. The study population is enriched by participants from racial and ethnic minority populations, underserved populations, and populations who experience poorer medical outcomes. All 10 eMERGE clinical sites conducted focus groups, interviews, and/or surveys to understand educational needs among key stakeholders-participants, providers, and/or study staff. Together, these studies highlighted the need for tools that address the perceived benefit/value of PRS, types of education/support needed, accessibility, and PRS-related knowledge and understanding. Based on findings from these preliminary studies, the network harmonized training initiatives and formal/informal educational resources. This paper summarizes eMERGE's collective approach to assessing educational needs and developing educational approaches for primary stakeholders. It discusses challenges encountered and solutions provided.

Returning integrated genomic risk and clinical recommendations: The eMERGE study

PURPOSE

Assessing the risk of common, complex diseases requires consideration of clinical risk factors as well as monogenic and polygenic risks, which in turn may be reflected in family history. Returning risks to individuals and providers may influence preventive care or use of prophylactic therapies for those individuals at high genetic risk.

METHODS

To enable integrated genetic risk assessment, the eMERGE (electronic MEdical Records and GEnomics) network is enrolling 25,000 diverse individuals in a prospective cohort study across 10 sites. The network developed methods to return cross-ancestry polygenic risk scores, monogenic risks, family history, and clinical risk assessments via a genome-informed risk assessment (GIRA) report and will assess uptake of care recommendations after return of results.

RESULTS

GIRAs include summary care recommendations for 11 conditions, education pages, and clinical laboratory reports. The return of high-risk GIRA to individuals and providers includes guidelines for care and lifestyle recommendations. Assembling the GIRA required infrastructure and workflows for ingesting and presenting content from multiple sources. Recruitment began in February 2022.

CONCLUSION

Return of a novel report for communicating monogenic, polygenic, and family history-based risk factors will inform the benefits of integrated genetic risk assessment for routine health care.

Human-centered development of an electronic health record-embedded, interactive information visualization in the emergency department using fast healthcare interoperability resources

OBJECTIVE

Develop and evaluate an interactive information visualization embedded within the electronic health record (EHR) by following human-centered design (HCD) processes and leveraging modern health information exchange standards.

MATERIALS AND METHODS

We applied an HCD process to develop a Fast Healthcare Interoperability Resources (FHIR) application that displays a patient's asthma history to clinicians in a pediatric emergency department. We performed a preimplementation comparative system evaluation to measure time on task, number of screens, information retrieval accuracy, cognitive load, user satisfaction, and perceived utility and usefulness. Application usage and system functionality were assessed using application logs and a postimplementation survey of end users.

RESULTS

Usability testing of the Asthma Timeline Application demonstrated a statistically significant reduction in time on task (P < .001), number of screens (P < .001), and cognitive load (P < .001) for clinicians when compared to base EHR functionality. Postimplementation evaluation demonstrated reliable functionality and high user satisfaction.

DISCUSSION

Following HCD processes to develop an application in the context of clinical operations/quality improvement is feasible. Our work also highlights the potential benefits and challenges associated with using internationally recognized data exchange standards as currently implemented.

CONCLUSION

Compared to standard EHR functionality, our visualization increased clinician efficiency when reviewing the charts of pediatric asthma patients. Application development efforts in an operational context should leverage existing health information exchange standards, such as FHIR, and evidence-based mixed methods approaches.

Development and dissemination of clinical decision support across institutions: standardization and sharing of refugee health screening modules

OBJECTIVES

We developed and piloted a process for sharing guideline-based clinical decision support (CDS) across institutions, using health screening of newly arrived refugees as a case example.

MATERIALS AND METHODS

We developed CDS to support care of newly arrived refugees through a systematic process including a needs assessment, a 2-phase cognitive task analysis, structured preimplementation testing, local implementation, and staged dissemination. We sought consensus from prospective users on CDS scope, applicable content, basic supported workflows, and final structure. We documented processes and developed sharable artifacts from each phase of development. We publically shared CDS artifacts through online dissemination platforms. We collected feedback and implementation data from implementation sites.

RESULTS

Responses from 19 organizations demonstrated a need for improved CDS for newly arrived refugee patients. A guided multicenter workflow analysis identified 2 main workflows used by organizations that would need to be supported by shared CDS. We developed CDS through an iterative design process, which was successfully disseminated to other sites using online dissemination repositories. Implementation sites had a small-to-modest analyst time commitment but reported a good match between CDS and workflow.

CONCLUSION

Sharing of CDS requires overcoming technical and workflow barriers. We used a guided multicenter workflow analysis and online dissemination repositories to create flexible CDS that has been adapted at 3 sites. Organizations looking to develop sharable CDS should consider evaluating the workflows of multiple institutions and collecting feedback on scope, design, and content in order to make a more generalizable product.

Formative Usability Testing Reduces Severe Blood Product Ordering Errors

BACKGROUND

Medical errors in blood product orders and administration are common, especially for pediatric patients. A failure modes and effects analysis in our health care system indicated high risk from the electronic blood ordering process.

OBJECTIVES

There are two objectives of this study as follows:(1) To describe differences in the design of the original blood product orders and order sets in the system (original design), new orders and order sets designed by expert committee (DEC), and a third-version developed through user-centered design (UCD).(2) To compare the number and type of ordering errors, task completion rates, time on task, and user preferences between the original design and that developed via UCD.

METHODS

A multidisciplinary expert committee proposed adjustments to existing blood product order sets resulting in the DEC order set. When that order set was tested with front-line users, persistent failure modes were detected, so orders and order sets were redesigned again via formative usability testing. Front-line users in their native clinical workspaces were observed ordering blood in realistic simulated scenarios using a think-aloud protocol. Iterative adjustments were made between participants. In summative testing, participants were randomized to use the original design or UCD for five simulated scenarios. We evaluated differences in ordering errors, time on task, and users' design preference with two-sample t-tests.

RESULTS

Formative usability testing with 27 providers from seven specialties led to 18 changes made to the DEC to produce the UCD. In summative testing, error-free task completion for the original design was 36%, which increased to 66% in UCD (30%, 95% confidence interval [CI]: 3.9-57%; p = 0.03). Time on task did not vary significantly.

CONCLUSION

UCD led to substantially different blood product orders and order sets than DEC. Users made fewer errors when ordering blood products for pediatric patients in simulated scenarios when using the UCD orders and order sets compared with the original design.

Identifying Electronic Health Record Usability And Safety Challenges In Pediatric Settings

Pediatric populations are uniquely vulnerable to the usability and safety challenges of electronic health records (EHRs), particularly those related to medication, yet little is known about the specific issues contributing to hazards. To understand specific usability issues and medication errors in the care of children, we analyzed 9,000 patient safety reports, made in the period 2012-17, from three different health care institutions that were likely related to EHR use. Of the 9,000 reports, 3,243 (36 percent) had a usability issue that contributed to the medication event, and 609 (18.8 percent) of the 3,243 might have resulted in patient harm. The general pattern of usability challenges and medication errors were the same across the three sites. The most common usability challenges were associated with system feedback and the visual display. The most common medication error was improper dosing.

A Heuristic Evaluation to Assess Use of After Visit Summaries for Supporting Continuity of Care

BACKGROUND

Outpatient providers often do not receive discharge summaries from acute care providers prior to follow-up visits. These outpatient providers may use the after-visit summaries (AVS) that are given to patients to obtain clinical information. It is unclear how effectively AVS support care coordination between clinicians.

OBJECTIVES

Goals for this effort include: (1) developing usability heuristics that may be applied both for assessment and to guide generation of medical documents in general, (2) conducting a heuristic evaluation to assess the use of AVS for communication between clinicians, and (3) providing recommendations for generating AVS that effectively support both patient/caregiver use and care coordination.

METHODS

We created a 17-item heuristic evaluation instrument for assessing usability of medical documents. Eight experts used the instrument to assess each of four simulated AVS. The simulations were created using examples from two hospitals and two pediatric patient cases developed by the National Institute of Standards and Technology.

RESULTS

Experts identified 224 unique usability problems ranging in severity from mild to catastrophic. Content issues (e.g., missing medical history, marital status of a 2-year-old) were rated as most severe, but widespread formatting and structural problems (e.g., inconsistent indentation, fonts, and headings; confusing ordering of information) were so distracting that they significantly reduced readers' ability to efficiently use the documents. Overall, issues in the AVS from Hospital 2 were more severe than those in the AVS from Hospital 1.

CONCLUSION

The new instrument allowed for quick, inexpensive evaluations of AVS. Usability issues such as unnecessary information, poor organization, missing information, and inconsistent formatting make it hard for patients, caregivers, and clinicians to use the AVS. The heuristics in the new instrument may be used as guidance to adapt electronic health record systems so that they generate more useful and usable medical documents.

Rapid Adjustment of Clinical Decision Support in Response to Updated Recommendations for Palivizumab Eligibility

Background: Palivizumab is effective at reducing hospitalizations due to respiratory syncytial virus among high-risk children, but is indicated for a small population. Identification of patients eligible to receive palivizumab is labor-intensive and error-prone. To support patient identification we developed Clinical Decision Support (CDS) based on published recommendations in 2012. This CDS was developed using a systematic process, which directly linked computer code to a recommendation’s narrative text. In 2014, updated recommendations were published, which changed several key criteria used to determine eligible patients.

Objective: Assess the effort required to update CDS in response to new palivizumab recommendations and identify factors that impacted these efforts.

Methods: We reviewed the updated American Academy of Pediatrics (AAP) policy statement from Aug 2014 and identified areas of divergence from the prior publication. We modified the CDS to account for each difference. We recorded time spent on each activity to approximate the total effort required to update the CDS.

Results: Of the 15 recommendations in the initial policy statement, 7 required updating. The CDS update was completed in 11 person-hours. Comparison of old and new recommendations was facilitated by the AAP policy statement structure and required 3 hours. Validation of the revised logic required 2 hours by a clinical domain expert. An informaticist required 3 hours to update and test the CDS. This included adding 24 lines and deleting 37 lines of code. Updating relevant data queries took an additional 3 hours and involved 10 edits.

Conclusion: We quickly adapted CDS in response to changes in recommendations for palivizumab administration. The consistent AAP policy statement structure and the link we developed between these statements and the CDS rules facilitated our efforts. We recommend that CDS implementers establish linkages between published narrative recommendations and their executable rules to facilitate maintenance efforts.

Citation: Michel J, Utidjian LH, Karavite D, Hogan A, Ramos MJ, Miller J, Shiffman RN, Grundmeier RW. Rapid adjustment of clinical decision support in response to updated recommendations for palivizumab eligibility. Appl Clin Inform 2017; 8: 581–592 https://doi.org/10.4338/ACI-2016-10-RA-0173

Adoption of a Portal for the Primary Care Management of Pediatric Asthma: A Mixed-Methods Implementation Study

Background

Patient portals may improve communication between families of children with asthma and their primary care providers and improve outcomes. However, the feasibility of using portals to collect patient-reported outcomes from families and the barriers and facilitators of portal implementation across diverse pediatric primary care settings have not been established.

Objective

We evaluated the feasibility of using a patient portal for pediatric asthma in primary care, its impact on management, and barriers and facilitators of implementation success.

Methods

We conducted a mixed-methods implementation study in 20 practices (11 states). Using the portal, parents of children with asthma aged 6-12 years completed monthly surveys to communicate treatment concerns, treatment goals, symptom control, medication use, and side effects. We used logistic regression to evaluate the association of portal use with child characteristics and changes to asthma management. Ten clinician focus groups and 22 semistructured parent interviews explored barriers and facilitators of use in the context of an evidence-based implementation framework.

Results

We invited 9133 families to enroll and 237 (2.59%) used the portal (range by practice, 0.6%-13.6%). Children of parents or guardians who used the portal were significantly more likely than nonusers to be aged 6-9 years (vs 10-12, P=.02), have mild or moderate/severe persistent asthma (P=.009 and P=.04), have a prescription of a controller medication (P<.001), and have private insurance (P=.002). Portal users with uncontrolled asthma had significantly more medication changes and primary care asthma visits after using the portal relative to the year earlier (increases of 14% and 16%, respectively). Qualitative results revealed the importance of practice organization (coordinated workflows) as well as family (asthma severity) and innovation (facilitated communication and ease of use) characteristics for implementation success.

Conclusions

Although use was associated with higher treatment engagement, our results suggest that achieving widespread portal adoption is unlikely in the short term. Implementation efforts should include workflow redesign and prioritize enrollment of symptomatic children.

ClinicalTrial

Clinicaltrials.gov NCT01966068; https://clinicaltrials.gov/ct2/show/NCT01966068 (Archived by WebCite at http://www.webcitation.org/6i9iSQkm3)

Clinical Decision Support and Palivizumab: A Means to Protect from Respiratory Syncytial Virus

BACKGROUND AND OBJECTIVES

Palivizumab can reduce hospitalizations due to respiratory syncytial virus (RSV), but many eligible infants fail to receive the full 5-dose series. The efficacy of clinical decision support (CDS) in fostering palivizumab receipt has not been studied. We sought a comprehensive solution for identifying eligible patients and addressing barriers to palivizumab administration.

METHODS

We developed workflow and CDS tools targeting patient identification and palivizumab administration. We randomized 10 practices to receive palivizumab-focused CDS and 10 to receive comprehensive CDS for premature infants in a 3-year longitudinal cluster-randomized trial with 2 baseline and 1 intervention RSV seasons.

RESULTS

There were 356 children eligible to receive palivizumab, with 194 in the palivizumab-focused group and 162 in the comprehensive CDS group. The proportion of doses administered to children in the palivizumab-focused intervention group increased from 68.4% and 65.5% in the two baseline seasons to 84.7% in the intervention season. In the comprehensive intervention group, proportions of doses administered declined during the baseline seasons (from 71.9% to 62.4%) with partial recovery to 67.9% during the intervention season. The palivizumab-focused group improved by 19.2 percentage points in the intervention season compared to the prior baseline season (p < 0.001), while the comprehensive intervention group only improved 5.5 percentage points (p = 0.288). The difference in change between study groups was significant (p = 0.05).

CONCLUSIONS

Workflow and CDS tools integrated in an EHR may increase the administration of palivizumab. The support focused on palivizumab, rather than comprehensive intervention, was more effective at improving palivizumab administration.

A survey of informatics approaches to whole-exome and whole-genome clinical reporting in the electronic health record

PURPOSE

Genome-scale clinical sequencing is being adopted more broadly in medical practice. The National Institutes of Health developed the Clinical Sequencing Exploratory Research (CSER) program to guide implementation and dissemination of best practices for the integration of sequencing into clinical care. This study describes and compares the state of the art of incorporating whole-exome and whole-genome sequencing results into the electronic health record, including approaches to decision support across the six current CSER sites.

METHODS

The CSER Medical Record Working Group collaboratively developed and completed an in-depth survey to assess the communication of genome-scale data into the electronic health record. We summarized commonalities and divergent approaches.

RESULTS

Despite common sequencing platform (Illumina) adoptions, there is a great diversity of approaches to annotation tools and workflow, as well as to report generation. At all sites, reports are human-readable structured documents available as passive decision support in the electronic health record. Active decision support is in early implementation at two sites.

CONCLUSION

The parallel efforts across CSER sites in the creation of systems for report generation and integration of reports into the electronic health record, as well as the lack of standardized approaches to interfacing with variant databases to create active clinical decision support, create opportunities for cross-site and vendor collaborations.

Effectiveness of decision support for families, clinicians, or both on HPV vaccine receipt

OBJECTIVE

To improve human papillomavirus (HPV) vaccination rates, we studied the effectiveness of targeting automated decision support to families, clinicians, or both.

METHODS

Twenty-two primary care practices were cluster-randomized to receive a 3-part clinician-focused intervention (education, electronic health record-based alerts, and audit and feedback) or none. Overall, 22, 486 girls aged 11 to 17 years due for HPV vaccine dose 1, 2, or 3 were randomly assigned within each practice to receive family-focused decision support with educational telephone calls. Randomization established 4 groups: family-focused, clinician-focused, combined, and no intervention. We measured decision support effectiveness by final vaccination rates and time to vaccine receipt, standardized for covariates and limited to those having received the previous dose for HPV #2 and 3. The 1-year study began in May 2010.

RESULTS

Final vaccination rates for HPV #1, 2, and 3 were 16%, 65%, and 63% among controls. The combined intervention increased vaccination rates by 9, 8, and 13 percentage points, respectively. The control group achieved 15% vaccination for HPV #1 and 50% vaccination for HPV #2 and 3 after 318, 178, and 215 days. The combined intervention significantly accelerated vaccination by 151, 68, and 93 days. The clinician-focused intervention was more effective than the family-focused intervention for HPV #1, but less effective for HPV #2 and 3.

CONCLUSIONS

A clinician-focused intervention was most effective for initiating the HPV vaccination series, whereas a family-focused intervention promoted completion. Decision support directed at both clinicians and families most effectively promotes HPV vaccine series receipt.

The implementation and acceptability of an HPV vaccination decision support system directed at both clinicians and families

We developed an electronic medical record (EMR)-based HPV vaccine decision support intervention targeting clinicians, (immunization alerts, education, and feedback) and families (phone reminders and referral to an educational website). Through telephone surveys completed by 162 parents of adolescent girls, we assessed the acceptability of the family-focused intervention and its effect on information-seeking behavior, communication, and HPV vaccine decision-making. The intervention was acceptable to parents and 46% remembered receiving the reminder call. Parents reported that the call prompted them to seek out information regarding the HPV vaccine, discuss the vaccine with friends and family, and reach a decision. Parents whose adolescent girls attended practices receiving the clinician-focused intervention were more likely to report that their clinician discussed the HPV vaccine at preventive visits. The results of this study demonstrate the acceptability and potential impact on clinical care of a comprehensive decision support system directed at both clinicians and families.

Reducing costs and improving outcomes of percutaneous coronary interventions

OBJECTIVE

To describe cost reduction and quality improvement efforts in our percutaneous coronary intervention (PCI) program and how risk adjustment was used to assess the effects of these changes.

STUDY DESIGN

Single center registry analysis.

PATIENTS AND METHODS

Data were collected on 2158 PCIs performed between July 1, 1994, and June 30, 1997. Of these, 1126 PCIs reflected care provided after implementation of competitive bidding for catheterization lab supplies, and efforts to reduce the use of postprocedure heparin and to implement early arterial sheaths removal (postbidding period). Hospital costs were estimated using a microcost accounting method. In-hospital mortality rates during the 2 time periods were compared using standardized mortality ratio estimated with a previously validated risk adjustment model for in-hospital mortality.

RESULTS

Compared with the prebidding period, the postbidding period was characterized by a significantly higher utilization of new technology (coronary stents and atherectomy devices 46% vs 25%; abciximab 19.1% vs 3.7, P<.01), and an overall increase in case complexity. Despite these changes, the average and median postbidding cost per case was dollars 1223 and dollars 1444 lower, respectively, than in the prebidding period. After adjustment for comorbidities, procedure variables, complications, and length of hospital stay, multivariate regression modeling identified the postbidding period as an independent predictor of lower hospital costs (P<.001) with an estimated adjusted cost savings of dollars 460. These cost savings were associated with trends toward a lower observed mortality rate, a higher predicted mortality rate, and a significantly lower standardized mortality ratio (SMR .71; 95% CI 0.48-0.9; P<.05).

CONCLUSION

Despite an increase in case complexity and utilization of new technology, cost reductions can be achieved through competitive bidding for supplies and modifications of periprocedure care. Risk adjustment appears to be a valid tool for assessing the effectiveness of these efforts independently from changes in case mix.

American College of Cardiology/American Heart Association preoperative assessment guidelines reduce resource utilization before aortic surgery

BACKGROUND

Methods used for evaluation of cardiac risk before noncardiac surgery vary widely. We evaluated the effect over time on practice and resource utilization of implementing the American College of Cardiology/American Heart Association Guidelines on Preoperative Risk Assessment.

METHODS

We compared 102 historical control patients who underwent elective abdominal aortic surgery (from January 1993 to December 1994) with 94 consecutive patients after guideline implementation (from July 1995 to December 1996) and 104 patients in a late after guideline implementation (from July 1, 1997, to September 30, 1998). Resource use (testing, revascularization, and costs) and outcomes (perioperative death and myocardial infarction) were examined. Patients with and without clinical markers of risk for perioperative cardiac complications were compared.

RESULTS

The use of preoperative stress testing (88% to 47%; P <.00001), cardiac catheterization (24% to 11%; P <.05), and coronary revascularization (25% to 2%; P <.00001) decreased between control and postguideline groups, respectively. These changes persisted in the late postguideline group. Mean preoperative evaluation costs also fell ($1087 versus $171; P <.0001). Outcomes of death (4% versus 3% versus 2%) and myocardial infarction (7% versus 3% versus 5%) were not significantly different between control, postguideline, and late postguideline groups, respectively. Stress test rates were similar for patients at low risk versus high risk in the historical control group (84% versus 91%; P =.29) but lower for patients at low risk after guideline implementation (31% versus 61%; P =.003).

CONCLUSION

Implementation of the American College of Cardiology/American Heart Association cardiac risk assessment guidelines appropriately reduced resource use and costs in patients who underwent elective aortic surgery without affecting outcomes. This effect was sustained 2 years after guideline implementation.

Development of a multicenter interventional cardiology database: the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2) experience

The technical challenges in the development of a quality-controlled registry of percutaneous coronary interventions (PCIs) are currently unknown. This article describes the authors' experience in the development of a regional, quality-controlled PCI registry. In 1996, 16 centers in Michigan were invited to participate in a multicenter PCI registry. Nine centers agreed to a pilot data collection and, as of July 2001, eight centers are still actively collecting data. An Oracle database was developed by the coordinating center. A common data collection form and a standard set of definitions were agreed on during several meetings. Data validity was insured through review of each form by a trained nurse, by automatic database diagnostic routines, and by site visits that included a review of the catheterization laboratory logs and a review of randomly selected charts. The average number of forms requiring query resolution was 33% in 1997 (range 7-76%), and it decreased to 5% in 1999 (range 1.4-10%). The most commonly queried variables were outcomes prior to discharge, lesion category, lesion complexity, date of birth, device used, gender, postprocedural percent stenosis, presence of left main disease, and MI date. Invalid dates, identification of the doctor, the presence of duplicate forms, and of duplicate outcomes were additional common queries generated by the internal diagnostic routines. In conclusion, the number of queries and diagnostic reports generated in the database suggests that the development of a quality-controlled PCI registry requires the institution of a careful diagnostic and data quality assessment system.

Acute myocardial infarction in the young--The University of Michigan experience

BACKGROUND

The purpose of this study was to assess frequency, risk factors, treatment, and complications of very young patients with acute myocardial infarction (MI) at the University of Michigan Medical Center (UMMC).

METHODS

From a database of 976 consecutive patients admitted to the UMMC with acute MI between 1995 and 1998, we compared care and outcomes of patients divided into 3 age categories: <46 years, 46-54 years, and >54 years. Risk factors, presenting symptoms, type of MI, management, complications, and hospital outcomes of the 3 groups were evaluated.

RESULTS

Young patients represented >10% of all patients with acute MI, and >25% of these individuals were women, a number considerably higher than seen in previous studies. This group of young patients was more likely to have Q-wave MI and risk factors such as family history and tobacco use and less likely to have a history of angina. Although all 3 groups received similar inpatient treatment, there was more attention paid to risk factor modification such as smoking cessation and referral to cardiac rehabilitation in younger individuals. Young patients had fewer in-hospital complications and a lower mortality rate.

CONCLUSIONS

At the University of Michigan, >1 in 10 with acute MI is <46 years old. Data suggest that current management and aggressive risk factor modification are quite good in this particular group, and overall the mortality rate is very low.

Implications of increased left ventricular mass index on in-hospital outcomes in patients undergoing aortic valve surgery

BACKGROUND

Increased left ventricular mass index has been shown to be associated with higher mortality in epidemiologic studies. However, the effect of increased left ventricular mass index on outcomes in patients undergoing aortic valve replacement is unknown.

METHODS

We studied 473 consecutive patients undergoing elective aortic valve replacement to assess the influence of left ventricular mass index on outcomes in patients having this procedure. Echocardiographic left ventricular dimensions were used to calculate left ventricular mass index (considered increased if >134 g/m(2) in male patients and >110 g/m(2) in female patients).

RESULTS

Left ventricular mass index was increased in 24% of patients undergoing aortic valve replacement. Postprocedural complications (respiratory failure, renal insufficiency, congestive heart failure, and atrial and ventricular arrhythmias), length of stay in the intensive care unit, and in-hospital mortality were increased in patients with increased left ventricular mass index. Multivariable analysis identified prior valve surgery (odds ratio, 4.3; 95% confidence interval, 1.2-15.7; P =.030), left ventricular ejection fraction (odds ratio, 1.07; 95% confidence interval, 1.01-1.14; P =.020), history of hypertension (odds ratio, 8.2; 95% confidence interval, 2.2-30.4; P =.002), history of liver disease (odds ratio, 50.4; 95% confidence interval, 4.2-609.0; P =.002), and increased left ventricular mass index (odds ratio, 38; 95% confidence interval, 9.3-154.1; P <.001) as independent predictors of in-hospital mortality. Furthermore, low output syndrome was identified as the most common mode of death (36%) after aortic valve replacement in patients with increased left ventricular mass index.

CONCLUSIONS

Increased left ventricular mass index is associated with increased adverse in-hospital clinical outcomes in patients undergoing aortic valve replacement. Although this finding warrants special modification in perioperative management, further studies are needed to address whether outcomes in asymptomatic patients with aortic valve disease could be improved by earlier aortic valve replacement before a significant increase in left ventricular mass index.

Comparison of artificial neural networks with logistic regression in prediction of in-hospital death after percutaneous transluminal coronary angioplasty

OBJECTIVES

Our objective was to compare artificial neural networks (ANNs) with logistic regression for prediction of in-hospital death after percutaneous transluminal coronary angioplasty and to assess the impact of guiding initial ANN variable selection with univariate analysis.

BACKGROUND

ANNs can detect complex patterns within data. Criticisms include the unpredictability of variable selection. They have not previously been applied to outcomes modeling for percutaneous coronary interventions.

METHODS

A database of consecutive (n = 3019) percutaneous transluminal coronary angioplasty procedures from an academic tertiary referral center between July 1994 and July 1997 was used. An ANN was developed for 38 variables (unguided model) (n = 1554). A second model was developed with predictors from an univariate analysis (guided model). Both were compared with a logistic regression model developed from the same database. Model validation was performed on independent data (n = 1465). Model predictive accuracy was assessed by the area under receiver operating characteristic curves. Goodness of fit was assessed with the Hosmer-Lemeshow statistic.

RESULTS

Sixty unguided and guided ANNs were developed. Predictive accuracy and model calibration for all models were similar for training data but were significantly better for logistic regression for independent validation data. Overestimation of event rate in higher risk patients accounted for the majority of discrepancy in model calibration for the ANNs. This difference was partially amended by guiding variable selection.

CONCLUSION

ANNs were able to model in-hospital death after percutaneous transluminal coronary angioplasty when guiding variable selection. However, performance was not better than traditional modeling techniques. Further investigations are needed to understand the impact of this methodology on outcomes analysis.

Usefulness of hypotension during dobutamine echocardiography in predicting perioperative cardiac events

This study was undertaken to determine the prognostic significance of hypotension induced during preoperative dobutamine stress echocardiography (DSE) before vascular and noncardiac thoracic surgery. Wall motion abnormality during DSE predicts perioperative risk. Although hypotension during DSE has not been shown to correlate with the presence or severity of coronary artery disease, its significance in perioperative risk assessment is unknown. We retrospectively studied 300 patients who had DSE within 6 months of noncardiac surgery. Perioperative events including death, myocardial infarction, ischemia, and arrhythmias were recorded. Odds ratios with 95% confidence intervals were used to examine the association between clinical and echocardiographic variables and perioperative events. A hypotensive response during DSE was seen in 85 patients (28%). Forty-eight patients (16%) had 54 perioperative complications including 4 cardiac-related deaths, 10 myocardial infarctions, 12 myocardial ischemic events, and 28 arrhythmias. Hypotension during DSE was predictive of the combined end point of perioperative cardiac mortality, myocardial infarction, and ischemia (odds ratio 4.04, 95% confidence interval 1.72 to 9.51). In a multivariate logistic regression model, hypotension during DSE remained a significant predictor (odds ratio 4.10, p<0.01). DSE-related hypotension was predictive of perioperative cardiac events and therefore may have a role in risk stratification before vascular or noncardiac thoracic surgery.

Validation of risk adjustment models for in-hospital percutaneous transluminal coronary angioplasty mortality on an independent data set

OBJECTIVES

We sought to validate recently proposed risk adjustment models for in-hospital percutaneous transluminal coronary angioplasty (PTCA) mortality on an independent data set of high risk patients undergoing PTCA.

BACKGROUND

Risk adjustment models for PTCA mortality have recently been reported, but external validation on independent data sets and on high risk patient groups is lacking.

METHODS

Between July 1, 1994 and June 1, 1996, 1,476 consecutive procedures were performed on a high risk patient group characterized by a high incidence of cardiogenic shock (3.3%) and acute myocardial infarction (14.3%). Predictors of in-hospital mortality were identified using multivariate logistic regression analysis. Two external models of in-hospital mortality, one developed by the Northern New England Cardiovascular Disease Study Group (model NNE) and the other by the Cleveland Clinic (model CC), were compared using receiver operating characteristic (ROC) curve analysis.

RESULTS

In this patient group, an overall in-hospital mortality rate of 3.4% was observed. Multivariate regression analysis identified risk factors for death in the hospital that were similar to the risk factors identified by the two external models. When fitted to the data set, both external models had an area under the ROC curve >0.85, indicating overall excellent model discrimination, and both models were accurate in predicting mortality in different patient subgroups. There was a trend toward a greater ability to predict mortality for model NNE as compared with model CC, but the difference was not significant.

CONCLUSIONS

Predictive models for PTCA mortality yield comparable results when applied to patient groups other than the one on which the original model was developed. The accuracy of the two models tested in adjusting for the relatively high mortality rate observed in this patient group supports their application in quality assessment or quality improvement efforts.

Validation in a community hospital setting of a clinical rule to predict preserved left ventricular ejection fraction in patients after myocardial infarction

BACKGROUND

A previous study showed that patients with previous myocardial infarction (MI) who meet 4 simple clinical and/or electrocardiographic criteria have a left ventricular ejection fraction (LVEF) of 40% or greater, with a positive predictive value of 98%. The objective of this study was to validate this clinical rule in the community hospital setting.

METHODS

Retrospective chart review in a 330-bed community hospital. Two hundred thirteen consecutive patients with MI were identified between June 1, 1993, and March 31, 1995. Left ventricular ejection fraction was predicted in a blinded fashion by means of the clinical rule before the actual LVEF test was reviewed.

RESULTS

We identified 213 patients admitted with the primary discharge diagnosis of acute MI. All patients met standard clinical and enzymatic definitions for acute MI and had at least 1 measure of LVEF, such as echocardiography, ventricular angiography, or gated blood pool scan. The clinical rule predicted that 83 patients (39.0%) would have an LVEF of 40% or greater. Of these 83 patients, 71 had an ejection fraction of 40% or greater, for a positive predictive value of 86%. Of the 12 patients who were incorrectly predicted to have a preserved LVEF, 6 (50%) had an index non-Q-wave anterior MI (P<.001). Reanalyzing the patient population with a fifth variable (anterior non-Q-wave MI) added to the original 4 variables increased the positive predictive value to 91%.

CONCLUSION

This simple clinical prediction rule has a positive predictive value of 86% when applied in the community hospital setting. Patients with anterior non-Q-wave MI may be 1 group in whom the rule is inaccurate, and expanding the clinical rule to 5 variables may increase the positive predictive value. When a technology-based assessment of left ventricular function is considered in patients after an MI, this prediction rule may allow for a more cost-effective patient selection, and as many as 40% of patients who have had acute MIs may require no testing at all.

Frequency, predictors, and appropriateness of blood transfusion after percutaneous coronary interventions

Increased awareness of the risks of blood-borne infections has recently led to profound changes in the practice of transfusion medicine. These changes include, among others, the development of guidelines by the American College of Physicians (ACP) for transfusion. Although the incidence and predictors of vascular complications of percutaneous interventions have been well defined, there are currently no data on frequency, risk factors, and appropriateness of blood transfusions. We performed a retrospective analysis of 628 consecutive percutaneous coronary revascularization procedures. Predictors of blood transfusion were identified using multivariate logistic regression analysis. Appropriateness of transfusions was determined using modified ACP guidelines. Transfusions were administered after 8.9% of interventions (56 of 628). Multivariate analysis identified age >70 years, female gender, procedure duration, coronary stenting, acute myocardial infarction, postprocedural use of heparin and intra-aortic balloon pump placement as independent predictors of blood transfusions (all p <0.05). According to the ACP guidelines, 36 of 56 patients (64%) received transfusions inappropriately. Transfusion reactions (fever) occurred in 10% of patients who received tranfusions appropriately and in 5% of patients who received tranfusions inappropriately. The estimated additional costs per procedure related to transfusions were $551 and $419, respectively. In conclusion, unnecessary transfusions were performed frequently after percutaneous coronary interventions. Application of available guidelines could reduce the number of unnecessary transfusions, thus avoiding exposure of patients to additional risks and reducing procedural costs.

Surgical delay for acute type A dissection with malperfusion

BACKGROUND

An acute type A aortic dissection is considered a surgical emergency. Review of the risk factors for a type A dissection showed that preoperative malperfusion was associated with a 22% (2/9) intraoperative mortality and an 89% (8/9) hospital mortality. Intraoperative deaths were secondary to pulmonary failure resulting from capillary leak; the remaining patients died of multiorgan failure resulting from reperfusion injury.

METHODS

The surgical delay approach was adopted for malperfused patients, and treatment in these patients included percutaneous reperfusion, with aortic fenestration and branch stenting where appropriate. Twenty patients had a type A dissection and malperfusion shown by pulsed-wave Doppler echocardiography, transesophageal echocardiography, or spiral computed tomographic scanning. Malperfusion was documented by angiography. After reperfusion, all patients' conditions were stabilized in the intensive care unit; intravenous beta-blockers were administered to decrease the maximum rate of increase of left ventricular pressure. Once patients completely recovered from the consequences of malperfusion, surgical repair was performed. Statistical comparison of the non-delay and delay groups was performed using Fisher's exact test and Student's t test. Multiple logistic regression analysis was used to establish independent predictors for mortality.

RESULTS

The mean delay to repair was 20 days (2 to 67 days). Four (31%) patients were discharged home and readmitted for operation. Three patients (15%) died preoperatively, 1 of retrograde dissection and rupture and 2 of reperfusion injury. Seventeen underwent surgical repair, with two deaths (12%); 15 (75%) were discharged, with an average follow-up of 16.8 months (p < 0.003). Delay was the only independent predictor of outcome.

CONCLUSIONS

Patients with an acute type A dissection and malperfusion should undergo percutaneous reperfusion, and surgical repair should be delayed until the reperfusion injury resolves.