Electronic Health Record Prompt to Improve Lung Cancer Screening in Primary Care

INTRODUCTION

Lung cancer is the leading cause of cancer death in the U.S. Combusted tobacco use, the primary risk factor, accounts for 90% of all lung cancers. Early detection of lung cancer improves survival, yet lung cancer screening rates are much lower than those of other cancer screening tests. Electronic health record (EHR) systems are an underutilized tool that could improve screening rates.

METHODS

This study was conducted in the Rutgers Robert Wood Johnson Medical Group, a university-affiliated network in New Brunswick, NJ. Two novel EHR workflow prompts were implemented on July 1, 2018. These prompts included fields to determine tobacco use and lung cancer screening eligibility and facilitated low-dose computed tomography ordering for eligible patients. The prompts were designed to improve tobacco use data entry, allowing for better lung cancer screening eligibility identification. Data were analyzed in 2022 retrospectively for the period July 1, 2017 to June 30, 2019. The analyses represented 48,704 total patient visits.

RESULTS

The adjusted odds of patient record completeness to determine eligibility for low-dose computed tomography (AOR=1.19, 95% CI=1.15, 1.23), eligibility for low-dose computed tomography (AOR=1.59, 95% CI=1.38, 1.82), and whether low-dose computed tomography was ordered (AOR=1.04, 95% CI=1.01, 1.07) all significantly increased after the electronic medical record prompts were implemented.

CONCLUSIONS

These findings show the utility and benefit of EHR prompts in primary care settings to increase identification for lung cancer screening eligibility as well as increased low-dose computed tomography ordering.

Time-of-day changes in physician clinical decision making: A retrospective study

Background

Time of day has been associated with variations in certain clinical practices such as cancer screening rates. In this study, we assessed how more general process measures of physician activity, particularly rates of diagnostic test ordering and diagnostic assessments, might be affected by time of day.

Methods

We conducted a retrospective chart review of 3,342 appointments by 20 attending physicians at five outpatient clinics, matching appointments by physician and comparing the average diagnostic tests ordered and average diagnoses assessed per appointment in the first hour of the day versus the last hour of the day. Statistical analyses used sign tests, two-sample t-tests, Wilcoxon tests, Kruskal Wallis tests, and multivariate linear regression.

Results

Examining physicians individually, four and six physicians, respectively, had statistically significant first- versus last-hour differences in the number of diagnostic tests ordered and number of diagnoses assessed per patient visit (p ≤ 0.04). As a group, 16 of 20 physicians ordered more tests on average in the first versus last hour (p = 0.012 for equal chance to order more in each time period). Substantial intra-clinic heterogeneity was found in both outcomes for four of five clinics (p < 0.01).

Conclusions

There is some statistical evidence on an individual and group level to support the presence of time-of-day effects on the number of diagnostic tests ordered per patient visit. These findings suggest that time of day may be a factor influencing fundamental physician behavior and processes. Notably, many physicians exhibited significant variation in the primary outcomes compared to same-specialty peers. Additional work is necessary to clarify temporal and inter-physician variation in the outcomes of interest.

Clinical research during the COVID-19 pandemic: The role of virtual visits and digital approaches

Clinical trials are a fundamental tool in evaluating the safety and efficacy of new drugs, medical devices, and health system interventions. Clinical trial visits generally involve eligibility assessment, enrollment, intervention administration, data collection, and follow-up, with many of these steps performed during face-to-face visits between participants and the investigative team. Social distancing, which emerged as one of the mainstay strategies for reducing the spread of SARS-CoV-2, has presented a challenge to the traditional model of clinical trial conduct, causing many research teams to halt all in-person contacts except for life-saving research. Nonetheless, clinical research has continued during the pandemic because study teams adapted quickly, turning to virtual visits and other similar methods to complete critical research activities. The purpose of this special communication is to document this rapid transition to virtual methodologies at Clinical and Translational Science Awards hubs and highlight important considerations for future development. Looking beyond the pandemic, we envision that a hybrid approach, which implements remote activities when feasible but also maintains in-person activities as necessary, will be adopted more widely for clinical trials. There will always be a need for in-person aspects of clinical research, but future study designs will need to incorporate remote capabilities.

Lessons learned from implementing service-oriented clinical decision support at four sites: A qualitative study

OBJECTIVE

To identify challenges, lessons learned and best practices for service-oriented clinical decision support, based on the results of the Clinical Decision Support Consortium, a multi-site study which developed, implemented and evaluated clinical decision support services in a diverse range of electronic health records.

METHODS

Ethnographic investigation using the rapid assessment process, a procedure for agile qualitative data collection and analysis, including clinical observation, system demonstrations and analysis and 91 interviews.

RESULTS

We identified challenges and lessons learned in eight dimensions: (1) hardware and software computing infrastructure, (2) clinical content, (3) human-computer interface, (4) people, (5) workflow and communication, (6) internal organizational policies, procedures, environment and culture, (7) external rules, regulations, and pressures and (8) system measurement and monitoring. Key challenges included performance issues (particularly related to data retrieval), differences in terminologies used across sites, workflow variability and the need for a legal framework.

DISCUSSION

Based on the challenges and lessons learned, we identified eight best practices for developers and implementers of service-oriented clinical decision support: (1) optimize performance, or make asynchronous calls, (2) be liberal in what you accept (particularly for terminology), (3) foster clinical transparency, (4) develop a legal framework, (5) support a flexible front-end, (6) dedicate human resources, (7) support peer-to-peer communication, (8) improve standards.

CONCLUSION

The Clinical Decision Support Consortium successfully developed a clinical decision support service and implemented it in four different electronic health records and four diverse clinical sites; however, the process was arduous. The lessons identified by the Consortium may be useful for other developers and implementers of clinical decision support services.

Primary and secondary prevention of cardiovascular disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

BACKGROUND

This guideline focuses on long-term administration of antithrombotic drugs designed for primary and secondary prevention of cardiovascular disease, including two new antiplatelet therapies.

METHODS

The methods of this guideline follow those described in Methodology for the Development of Antithrombotic Therapy and Prevention of Thrombosis Guidelines: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines in this supplement.

RESULTS

We present 23 recommendations for pertinent clinical questions. For primary prevention of cardiovascular disease, we suggest low-dose aspirin (75-100 mg/d) in patients aged > 50 years over no aspirin therapy (Grade 2B). For patients with established coronary artery disease, defined as patients 1-year post-acute coronary syndrome, with prior revascularization, coronary stenoses > 50% by coronary angiogram, and/or evidence for cardiac ischemia on diagnostic testing, we recommend long-term low-dose aspirin or clopidogrel (75 mg/d) (Grade 1A). For patients with acute coronary syndromes who undergo percutaneous coronary intervention (PCI) with stent placement, we recommend for the first year dual antiplatelet therapy with low-dose aspirin in combination with ticagrelor 90 mg bid, clopidogrel 75 mg/d, or prasugrel 10 mg/d over single antiplatelet therapy (Grade 1B). For patients undergoing elective PCI with stent placement, we recommend aspirin (75-325 mg/d) and clopidogrel for a minimum duration of 1 month (bare-metal stents) or 3 to 6 months (drug-eluting stents) (Grade 1A). We suggest continuing low-dose aspirin plus clopidogrel for 12 months for all stents (Grade 2C). Thereafter, we recommend single antiplatelet therapy over continuation of dual antiplatelet therapy (Grade 1B).

CONCLUSIONS

Recommendations continue to favor single antiplatelet therapy for patients with established coronary artery disease. For patients with acute coronary syndromes or undergoing elective PCI with stent placement, dual antiplatelet therapy for up to 1 year is warranted.

Antithrombotic and thrombolytic therapy for ischemic stroke: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines

OBJECTIVES

This article provides recommendations on the use of antithrombotic therapy in patients with stroke or transient ischemic attack (TIA).

METHODS

We generated treatment recommendations (Grade 1) and suggestions (Grade 2) based on high (A), moderate (B), and low (C) quality evidence.

RESULTS

In patients with acute ischemic stroke, we recommend IV recombinant tissue plasminogen activator (r-tPA) if treatment can be initiated within 3 h (Grade 1A) or 4.5 h (Grade 2C) of symptom onset; we suggest intraarterial r-tPA in patients ineligible for IV tPA if treatment can be initiated within 6 h (Grade 2C); we suggest against the use of mechanical thrombectomy (Grade 2C) although carefully selected patients may choose this intervention; and we recommend early aspirin therapy at a dose of 160 to 325 mg (Grade 1A). In patients with acute stroke and restricted mobility, we suggest the use of prophylactic-dose heparin or intermittent pneumatic compression devices (Grade 2B) and suggest against the use of elastic compression stockings (Grade 2B). In patients with a history of noncardioembolic ischemic stroke or TIA, we recommend long-term treatment with aspirin (75-100 mg once daily), clopidogrel (75 mg once daily), aspirin/extended release dipyridamole (25 mg/200 mg bid), or cilostazol (100 mg bid) over no antiplatelet therapy (Grade 1A), oral anticoagulants (Grade 1B), the combination of clopidogrel plus aspirin (Grade 1B), or triflusal (Grade 2B). Of the recommended antiplatelet regimens, we suggest clopidogrel or aspirin/extended-release dipyridamole over aspirin (Grade 2B) or cilostazol (Grade 2C). In patients with a history of stroke or TIA and atrial fibrillation we recommend oral anticoagulation over no antithrombotic therapy, aspirin, and combination therapy with aspirin and clopidogrel (Grade 1B).

CONCLUSIONS

These recommendations can help clinicians make evidence-based treatment decisions with their patients who have had strokes.

Optimal medical therapy is superior to transplantation for the treatment of class I, II, and III heart failure: a decision analytic approach

BACKGROUND

The survival benefit of heart transplantation (HT) compared with optimal medical therapy (OMT) has never been tested.

METHODS AND RESULTS

We created a decision analytic model that simulates a randomized clinical trial of OMT versus HT for each New York Heart Association (NYHA) class. The simulation calculates average life expectancy. The following assumptions were made for OMT annual mortality: class I no excess mortality from HF; class II and III based on MERIT-HF are 5.3% and 8.1%. Class IV is 12.8%, based on COPERNICUS. HT mortality rates were based on survival curves for HT 1982 to 2001. For classes I, II, and III, OMT demonstrated a life expectancy gain of 113 months (232+/-2.2 versus 119+/-2.1), 38 months (152+/-2.1 versus 114+/-2.1), and 6 months (117+/-1.8 versus 111+/-2.2), respectively, over HT. Class IV favored HT with a life expectancy gain of 26 months (107+/-2.1 versus 81+/-1.4) over OMT. Sensitivity analysis revealed if improvement in OMT decreased mortality by 38% for class IV patients, OMT and HT would have equivalent life expectancies. If improvement in HT resulted in a 7% increase in post-HT survival, OMT and HT would be equivalent for class III patients. If improvement in HT resulted in a 30% increase in post-HT survival, OMT and HT would be equivalent for class II patients.

CONCLUSIONS

Our model predicts that currently, OMT is superior to HT for classes I, II, and III, but HT is superior for class IV. However, future advances in OMT or HT may change the relative benefits of these treatment modalities.

Prediction in annotation based guideline encoding

The encoding of clinical practice guidelines into machine operable representations poses numerous challenges and will require considerable human intervention for the foreseeable future. To assist and potentially speed up this process, we have developed an incremental approach to guideline encoding which begins with the annotation of the original guideline text using markup techniques. A modular and flexible sequence of subtasks results in increasingly inter-operable representations while maintaining the connections to all prior source representations and supporting knowledge. To reduce the encoding bottleneck we also employ a number of machine-assisted learning and prediction techniques within a knowledge-based software environment. Promising results with a straightforward incremental learning algorithm illustrate the feasibility of such an approach.

Computer-interpretable clinical practice guidelines. Where are we and where are we going ?

OBJECTIVES

To provide a comprehensive overview of computer-interpretable guideline (CIG) systems aimed at non-experts. The overview includes the history of efforts to develop CIGs, features of and relationships among current major CIG systems, current status of standards developments pertinent to CIGs and identification of unsolved problems and needs for future research.

METHODS

Literature review based on PubMed, AMIA conference proceedings and key references from publications identified. Search terms included practice guidelines, decision support, controlled vocabulary and medical record systems. Papers were reviewed by both authors and summarized narratively.

RESULTS

There is a consensus that guideline delivery systems must be integrated with electronic health records (EHRs) to be most effective. Several evolving CIG formalisms have in common, use of a task network model. There is currently no dominant CIG system. The major challenge in development of interoperable CIGs, is agreement on a standard controlled vocabulary. Such standards are under development, but not widely used, particularly in commercial EHR systems. The Virtual Medical Record (VMR) concept has been proposed as a standard that would serve as an intermediary between guideline vocabulary and that used in EHR implementation.

CONCLUSION

CIG systems are in a state of evolution. Standards efforts promise to improve interoperability without compromising innovation. The VMR concept can assist guideline development even before clinical systems routinely adhere to standards. Frontiers for future work include using the principles learned by computer implementation of guidelines to improve the guideline development process and evaluation methods that isolate the effects of specific CIG features.

Cost-effectiveness and contraceptive effectiveness of the transdermal contraceptive patch

OBJECTIVE

The purpose of this study was to examine implications of increased perfect use on the cost-effectiveness of the contraceptive patch compared with combination oral contraceptives (COCs).

STUDY DESIGN

This study compared the patch with low-estrogen-dose COCs. It assumes that the risks of developing a medical condition during use are the same for both the patch and COCs. Differences in net cost and pregnancies avoided during use were modeled. With the use of a pharmacoeconomic model, both methods were compared with a hypothetical reference case of contraception nonuse. The base-case model considered women, ages 15 to 50, in average health in a long-term, mutually monogamous, heterosexual relationship.

RESULTS

The base-case analysis showed that use of the patch resulted in a savings of 249 US dollars and 0.03 pregnancies per woman over 2 years compared with COCs.

CONCLUSION

This analysis demonstrated that patch use would be cost saving compared with COC use, resulting in a net avoidance of pregnancy for this population. The cost savings are attributed to reduced costs of pregnancy.

Vocabulary requirements for implementing clinical guidelines in an electronic medical record: a case study

As part of a larger effort to automate guidelines we determined the number and types of clinical variables required to implement two complex clinical guidelines and the adequacy of the electronic medical record (EMR) to capture them. 178 unique variables were required by both guidelines. Variables were classified as simple (existing observation terms in the EMR), calculated (transformations of simple variables), and complex (requiring multiple simple variables and logical rules for combining them). Many variables are unlikely to be instantiated in an EMR without focused efforts to collect them. In addition, many variables required knowledge that was neither provided in the guideline nor referenced. We conclude that, although the EMR contains the necessary variables to implement these guidelines, successful automated implementation requires unambiguous definition of required terms, incorporation of additional knowledge not provided in the guideline and modification of workflow to collect variables not normally captured in routine clinical care.

Costs and net health effects of contraceptive methods

BACKGROUND

Pregnancy and contraceptive methods both have important health effects that include risks and benefits. The net impact of contraception on women's health has not been reported previously.

STUDY DESIGN

This is a cost-utility analysis using a Markov model evaluated by Monte Carlo simulation using the societal perspective for costs. The analysis compared 13 methods of contraception to nonuse of contraception with respect to healthcare costs and quality-adjusted life years (QALYs). Discounting was applied for future costs and health effects. The base-case analysis applies to women of average health and fertility, ranging from 15 to 50 years of age, who are sexually active in a mutually monogamous relationship; smoking rates observed in women of reproductive age were used. Sensitivity analysis extended the analysis to nonmonogamous status and smoking status.

RESULTS

Compared with use of no contraception, contraceptive methods of all types result in substantial cost savings over 2 years, ranging from US$5907 per woman for tubal sterilization to US$9936 for vasectomy and health gains ranging from 0.088 QALYs for diaphragm to 0.147 QALYs for depot medroxyprogesterone acetate. Compared with nonuse, even with a time horizon as short as 1 year, use of any method other than sterilization results in financial savings and health gains. Most of the financial savings and health gains were due to contraceptive effects. In a population of patients, even modest increases in the use of the most effective methods result in financial savings and health gains.

CONCLUSIONS

Every method of contraception dominates nonuse in most clinical settings. Increasing the use of more effective methods even modestly at the expense of less effective methods will improve health and reduce costs. Methods that require action by the user less frequently than daily are both less costly and more effective than methods requiring action on a daily basis.

Decision analysis: a basic overview for the pediatric surgeon

Decision making in medicine requires choosing the option that best maximizes benefit while minimizing risk and cost. Even though uncertainty is an inherent feature of any clinical issue, clinicians and policy makers frequently are required to evaluate the best evidence and make therapeutic or policy decisions based on that evidence. Decision analysis is a quantitative approach to decision making under conditions of uncertainty that can be applied to specific types of clinical problems. This method disaggregates a complex clinical problem into its most important components that then can be understood more easily and analyzed quantitatively. Decision analysis has many potential applications in medicine and can be applied to solve specific clinical problems, analyze health care costs, or develop health care policies. In this review, the basic methods for constructing and analyzing decision analyses will be presented, and specific applications of this method to pediatric surgery will be discussed.

Health economics of contraception

Health economic analyses evaluate interventions from the standpoint of effectiveness and cost, important considerations in today's healthcare environment. Use of analytic cost-effectiveness approaches allows decision-makers to evaluate interventions in a more rational manner. With the use of sensitivity techniques, various scenarios can be evaluated, particularly as they relate to questions that have significant uncertainty. To date, there has been limited application of these methods to the area of contraception.

Is the societal perspective in cost-effectiveness analysis useful for decision makers?

BACKGROUND

The U.S. Public Health Service's Panel on Cost-Effectiveness in Health and Medicine recommends that cost-effectiveness analyses (CEAs) intended to help allocate health resources in the public interest include a reference case analysis, conducted from the perspective of society as a whole. This perspective requires that an analysis measure all health effects and changes in resource use caused by an intervention. VALUE OF SOCIETAL CEAs: Tensions unavoidably arise among the parties to medical decisions--patients, their families and friends, clinicians, and third-party payers. One common approach to handling these tensions is to ignore some of them, to "solve" them by pretending they do not or should not exist. Patients do this when they demand the best care for themselves without regard to the cost to others, payers when they make coverage decisions that shift costs ot others. But by estimating all gains and losses, calculations that reflect the safety, effectiveness, and side effects of an intervention as well as its costs, societal CEAs can help resolve those tensions and provide the basis for decisions that are fair to all parties, an agenda for negotiating such decisions, and information essential for designing compensation and incentives to support them. MAKING BETTER USE OF SOCIETAL CEAs: Use of the societal perspective asks that all parties be aware of and consider the interests of others. Some process or procedure needs to be developed for presenting CEA information to the parties most likely to be affected by decisions, soliciting their views, and negotiating an acceptable decision. This process could be used by government decision makers or by managed care organizations, professional societies, or payers.

The will to live among HIV-infected patients

BACKGROUND

Patients infected with HIV value both longevity and health.

OBJECTIVE

To understand how HIV-infected patients value their health.

DESIGN

Interview study.

SETTING

Regional treatment center for HIV.

PATIENTS

51 patients with HIV infection.

MEASUREMENTS

Life-satisfaction, health rating, time-tradeoff, and standard-gamble scores.

RESULTS

Of the 51 patients, 49% (95% CI, 35% to 63%) said that their life was better currently than it was before they contracted HIV infection; only 29% said that life was currently worse. The mean (+/- SD) time-tradeoff score was 0.95 +/- 0.10, indicating that, on average, patients would give up no more than 5% of their remaining life expectancy in their current state of health in exchange for a shorter but healthy life. The average health rating score was 71.0 +/- 18.7 on a scale of 0 to 100, and the average standard-gamble score was 0.80 +/- 0.27. Factors contributing to life satisfaction and time-tradeoff scores included spirituality and having children.

CONCLUSION

Many patients with HIV have a strong will to live, and many feel that life with HIV is better than it was before they became infected.

The cost-effectiveness of autologous transfusion revisited: implications of an increased risk of bacterial infection with allogeneic transfusion

BACKGROUND

Previous analyses have found autologous transfusion to be very expensive but have not considered avoidance of postoperative bacterial infections as one of its benefits.

STUDY DESIGN AND METHODS

A cost-utility analysis using a Markov cohort simulation model compared autologous blood transfusion to allogeneic transfusion in a hypothetical cohort of patients undergoing elective total hip replacement with respect to discounted quality-adjusted life years (QALYs) and health-care system costs.

RESULTS

Assuming a base case rate of serious infection of 3.7 percent, a relative risk of infection of 1.85, and additional costs of $12,980 per infection, autologous transfusion has a cost-effectiveness of $2,470 per QALY. If the relative risk of bacterial infection following allogeneic transfusion exceeds 1.1, the cost-effectiveness of autologous transfusion is less than $50,000 per QALY and if the relative risk exceeds 2.4, autologous transfusion is dominant, resulting in both lower costs and greater QALYs. If there were no increased risk of transfusion, the cost-effectiveness of autologous transfusion would be $3,400,000 per QALY.

CONCLUSIONS

If there is only a modest increase in the risk of bacterial infection following allogeneic transfusion, autologous transfusion would result in improved outcomes at a cost of less than $50,000 per QALY. Autologous transfusion would be dominant above a relative risk of infection that is within the range of values observed in randomized controlled trials. However, if there is no increased risk of bacterial infection, autologous transfusion would be a very expensive strategy. Until more definitive data are available on the magnitude and costs of this risk, we advise against prematurely closing the debate about the cost-effectiveness of autologous transfusion.

Risk of bacterial infection associated with allogeneic blood transfusion among patients undergoing hip fracture repair

BACKGROUND

The relationship between allogeneic blood transfusion and bacterial infection remains uncertain. An increased risk of bacterial infection would represent the most important risk of allogeneic transfusion, because viral disease transmission has become so rare.

STUDY DESIGN AND METHODS

A retrospective cohort study of 9598 consecutive hip fracture patients at least 60 years old who underwent surgical repair was performed. The primary outcome was serious bacterial infection, defined as bacteremia, pneumonia, deep wound infection, or septic arthritis or osteomyelitis. Secondary outcomes included two individual infections, pneumonia and urinary tract infection (UTI), and the cost of infection. Hospital cost of infection was assessed by linking the study population to Medicare data.

RESULTS

Fifty-eight percent of patients received at least one transfusion. Serious bacterial infection occurred in 437 patients (4.6%); 28.8 percent of this group died during the hospital stay. Pneumonia occurred in 361 patients (3.8%) and UTI occurred in 1157 patients (12.1%). The adjusted risk of serious bacterial infection associated with transfusion was 1.35 (95% CI, 1.10-1.66). The adjusted risk for pneumonia was 1.52 (95% CI, 1.21-1.91), and that for UTI was 1.03 (95% CI, 0.91-1.17). A dose-response relationship was present for serious bacterial infection (p = 0.001) and pneumonia (p = 0.001). The cost of hospitalization was $14,000 greater for patients with serious infection than for patients without infection.

CONCLUSION

Blood transfusion is associated with a 35-percent greater risk of serious bacterial infection and a 52-percent greater risk of pneumonia. Postoperative infections are costly. The risk of bacterial infection may be the most common life-threatening adverse effect of allogeneic blood transfusion.

Distributed decision support using a web-based interface: prevention of sudden cardiac death

Although decision models can provide a formal foundation for guideline development and clinical decision support, their widespread use is often limited by the lack of platform-independent software that geographically dispersed users can access and use easily without extensive training. To address these limitations the authors developed a World Wide Web-based interface for previously developed decision models. They describe the use and functionality of the interface using a decision model that evaluates the cost-effectiveness of strategies for preventing sudden cardiac death. The system allows an analyst to use a web browser to interact with the decision model and to change the values of input variables within pre-specified ranges, to specify sensitivity or threshold analyses, to evaluate the decision model, and to view the results generated dynamically. The web site also provides linkages to an explanation of the model, and evidence tables for input variables. The system demonstrates a method for providing distributed decision support to remote users such as guideline developers, decision analysts, and potentially practicing physicians. The web interface provides platform-independent and almost universal access to a decision model. This approach can make distributed decision support both practical and economical, and has the potential to increase the usefulness of decision models by enabling a broader audience to incorporate systematic analyses into both policy and clinical decisions.

Toward a peer review process for medical decision analysis models

This paper presents a framework for a peer review process for medical decision analysis models. This framework is based on the collective experience of the members of the Inter-PORT Decision Modeling work group, a team of decision analysis experts comprising members from each of the Patient Outcomes Research Teams (PORTs), sponsored by the Agency for Health Care Policy and Research. Important general principles of correct model structure include choice of model type, perspective of the analysis, choice of utility scheme and identification of strategies that should be included and events that should be modeled. In addition, a set of rules for correct decision model structure may help to identify common errors. Although not necessarily exhaustive, this scheme provides an approach by which a reviewer may judge the adequacy of a decision model presented for publication or as the basis for a health policy.

Does physician performance explain interspecialty differences in malpractice claim rates?

Physician specialty has been consistently associated with different malpractice claim rates, with neurosurgery, orthopedics, and obstetrics and gynecology having the highest rates. Whether these differences reflect physician performance or other aspects of patient care that are unique for each specialty is unclear. A retrospective cohort study was performed including 12,829 physicians involved in 8,221 closed cases from 1977 to 1991. For each case an assessment was made whether the plaintiff received an award and whether the physician care of the patient was indefensible. An award was made to the plaintiff in 42% of cases. Physician care was considered indefensible in 23% of the cases. The specialties with the highest award rates were anesthesiology (58.3%), obstetrics and gynecology (47.5%), and radiology (43.0%). Neurosurgery had the lowest rate of award (30.2%). The specialties with the highest indefensibility rates were radiology (36%), obstetrics and gynecology (27%), and anesthesiology (27%). Neurosurgery had the lowest indefensibility rate (10%). These results were unchanged after controlling for physician age, degree, site of training, certification status, and severity of patient injury. In conclusion, differences in award rate and indefensibility are present, but not large enough to explain the large variation in specialty claim rates. This suggests that the variation in malpractice rates results from factors other than a meaningful difference in physician performance.

An architecture for knowledge-based construction of decision models

Clinical application of decision analysis has been limited by unfamiliarity of clinicians with the technique, large data requirements, and the length of time needed to construct models. In order to make decision modeling more accessible to clinicians, the authors developed a computer program to construct decision models automatically. The system contains two separate knowledge bases. One contains frames encoding knowledge of the medical domain, the evaluation of pulmonary disease in patients infected with the human immunodeficiency virus (HIV). The other contains rules of correct decision model construction that guide the selection of items from the domain knowledge base and their insertion into the decision model. The system can create either a tree or an influence diagram that satisfies previously published critiquing rules. The system has the potential to enable novices to construct useful decision models and to provide individualized decision-analytic advice to clinicians in real time.

Markov models in medical decision making: a practical guide

Markov models are useful when a decision problem involves risk that is continuous over time, when the timing of events is important, and when important events may happen more than once. Representing such clinical settings with conventional decision trees is difficult and may require unrealistic simplifying assumptions. Markov models assume that a patient is always in one of a finite number of discrete health states, called Markov states. All events are represented as transitions from one state to another. A Markov model may be evaluated by matrix algebra, as a cohort simulation, or as a Monte Carlo simulation. A newer representation of Markov models, the Markov-cycle tree, uses a tree representation of clinical events and may be evaluated either as a cohort simulation or as a Monte Carlo simulation. The ability of the Markov model to represent repetitive events and the time dependence of both probabilities and utilities allows for more accurate representation of clinical settings that involve these issues.

Who should be screened for HIV infection? A cost-effectiveness analysis

BACKGROUND

The advent of effective prophylactic treatments for asymptomatic persons infected with human immunodeficiency virus has led to interest in widespread screening programs. However, the costs of screening programs and therapy are high, and the prevalence of infection above which screening becomes an appropriate use of scarce health care dollars remains undetermined.

METHODS

To examine the cost-effectiveness of screening in populations with differing prevalences of infection, we developed a Markov model to compare costs and life expectancy for two strategies: (1) screening and prophylactic treatment for infected persons who have or who develop low CD4+ (T4) cell counts, and (2) no screening. Based on studies in the literature, we estimated the prevalence of HIV infection, the rate of T4-cell loss, the rates of developing the acquired immunodeficiency syndrome and Pneumocystis pneumonia stratified by T4 cell counts, the life expectancy with the acquired immunodeficiency syndrome, the efficacy of prophylactic therapies, and costs.

RESULTS

In populations with a prevalence of infection more than 5%, which includes known risk groups, screening costs less than $11,000 per life-year gained. In populations with a prevalence as low as 0.15%, screening costs only $29,000 per life-year gained. Even when the efficacy of zidovudine is assumed to be limited to 3 years, screening still costs less than $40,000 per life-year gained in populations with a prevalence of 0.5% or greater. However, in populations with a very low prevalence of infection (two to 10/100,000), such as members of the general population without reported risk factors, screening costs rise to between $290,000 and $1,277,400 per life-year gained.

CONCLUSIONS

When considering only direct medical benefits, screening for asymptomatic human immunodeficiency virus infection in the general population, without regard to reported risk factors or seroprevalence data, would be expensive. In populations with a prevalence of infection of 0.5% or greater, however, the cost-effectiveness of screening falls within the range of currently accepted medical practices. These results suggest that screening be offered routinely to all persons in defined populations, such as persons receiving care at hospitals or clinics, or residing in geographic areas, where the seroprevalence is 0.5% or more, and underscore the need to conduct seroprevalence studies to identify such populations.

Comparative cost-effectiveness analysis of theophylline and ipratropium bromide in chronic obstructive pulmonary disease. A three-center study

The charts of 311 patients receiving theophylline (T) and 289 patients receiving ipratropium bromide (IB) for COPD were reviewed to determine the total costs and cost-effectiveness of these 2 agents in 3 different health-care settings. A direct cost-accounting method assessed cost, and a Markov decision-analysis model calculated cost-effectiveness. Costs to treat toxic effects were greater for T versus IB. The types and incidences of toxic effects, by drug, were similar among the three centers. Overall costs for T were $121.40 per patient per therapy-month versus $84.56 per patient per therapy-month for IB, as determined by the cost-accounting method. The marginal cost was $366 for T over IB when extrapolated over 1 year using the Markov model. The Markov model also predicted that patients receiving IB had a greater number of complication-free therapy-months (measurement of effectiveness) than patients receiving T. We conclude that treatment with IB was less costly and more cost-effective than T.

Cost-effectiveness of the implantable cardioverter-defibrillator: effect of improved battery life and comparison with amiodarone therapy

The implantable cardioverter-defibrillator (ICD) greatly reduces the incidence of sudden cardiac death among patients with recurrent sustained ventricular tachycardia and fibrillation who do not respond to conventional antiarrhythmic therapy. A cost-effectiveness analysis was performed, comparing the ICD, amiodarone and conventional agents. Actual variable costs of hospitalization and follow-up care were used for 21 ICD- and 43 amiodarone-treated patients. Life expectancy and total variable costs were predicted with use of a Markov decision analytic model. Clinical event rates and probabilities were based on published reports or expert opinion. Life expectancy with an ICD (6.1 years) was 50% greater than that associated with treatment with amiodarone (3.9 years) and 2.5 times that associated with conventional treatment (2.5 years). Assuming replacement every 24 months, ICD lifetime treatment costs (in 1989 dollars) for a 55-year old patient are expected to be $89,600 compared with $24,800 for amiodarone and $16,100 for conventional therapy, yielding a marginal cost/effectiveness ratio for ICD versus amiodarone therapy of 1f429,200/year of life saved, which is comparable to that of other accepted medical treatments. If technologic improvements extend average battery life to 36 months, the marginal cost/effectiveness ratio would be $21,800/year of life saved, and at 96 months it would be $13,800/year of life saved. Patient age at implantation did not significantly affect these results. If quality of life on amiodarone therapy is 30% lower than that with the ICD, the marginal cost/effectiveness ratio decreases by 35%. If the quality of life for patients receiving drugs is 40% lower than that of patients treated with an ICD, use of the defibrillator becomes the dominant strategy.

The risks of blood transfusion: the relative influence of acquired immunodeficiency syndrome and non-A, non-B hepatitis

PURPOSE

The acquired immunodeficiency syndrome epidemic has greatly increased concern about the risk of blood transfusion. Many transfusions are now autologous, and when these are not available, both physicians and patients are more likely to question the advisability of transfusion. We evaluate the risk of preoperative blood transfusion and the contribution of human immunodeficiency virus (HIV) infection to that risk.

METHODS

We used decision analysis to characterize the risk associated with HIV infection in days of life lost. The contributions to risk of acute transfusion reaction, hepatitis B, and non-A, non-B hepatitis are also estimated. Sensitivity analyses show the implications for transfusion risk of recent information about HIV infection in the blood supply and a new test for hepatitis C.

RESULTS

The analysis shows that the contribution of HIV infection to the risk of death from transfusion, expressed in days of life expectancy lost, has become extremely small over the last several years. Currently, HIV infection accounts for less than 1% of the risk of death, while non-A, non-B hepatitis accounts for 97% to 98%. Further reductions in the risk of HIV infection, even to zero, will make relatively little difference in the safety of transfusion. The analysis also shows that the remaining risk from transfusion should decrease sharply, by more than two thirds, with the adoption of the test for hepatitis C.

CONCLUSIONS

Efforts to improve the safety of blood should focus on reducing the risk of non-A, non-B hepatitis. The remaining risk of HIV infection is very small.

Myocardial revascularization for chronic stable angina. Analysis of the role of percutaneous transluminal coronary angioplasty based on data available in 1989

No prospective, randomized clinical trial comparing coronary artery bypass grafting, percutaneous transluminal coronary angioplasty, and conservative therapy has been reported. To address when revascularization is indicated, we constructed a decision analytic model. Our model incorporates procedure-related mortality and morbidity, coronary artery disease-related mortality, and the benefit of revascularization. We determined the quality-adjusted life expectancy and expected costs for each strategy. Our model suggests that angioplasty is a reasonable alternative to bypass surgery in patients with favorable lesions if angioplasty would provide a comparable degree of revascularization. Our model predicts that patients treated with angioplasty will have more revascularization procedures than will patients treated with bypass surgery but predicts that both treatments will cost the same over the typical patient's lifetime. In many patients with severe angina or documented ischemia, angioplasty is indicated for stenosis of a single artery. In patients with two vessel disease that is amenable to angioplasty, angioplasty may be a reasonable alternative to bypass surgery. Even in patients whose three vessel disease can be completely revascularized by angioplasty, bypass surgery, although relatively expensive, is slightly better than angioplasty. In patients with three vessel disease and comorbidities that increase operative risk, angioplasty may be a reasonable alternative to either bypass surgery or medical therapy.

Bone marrow donor registries: the relation between registry size and probability of finding complete and partial matches

In a registry of volunteer bone marrow donors, the relation between registry size and probability of finding an exact or partial match for a random recipient cannot be theoretically derived because it depends on specifics of the human leukocyte antigen (HLA) haplotype frequencies in the donor and recipient populations. The relation must be explicitly calculated using empirically determined HLA haplotype frequency data for all possible pairings between a donor and a recipient population. This report describes a general solution to this problem. The method shows that the relation of the probability of matching to registry size is sigmoidal, with small increases in probability at the extremes of registry size and a middle range of registry size within which the probability of matching increases most sharply. This range determines the approximate size of the most cost-effective registry. In addition, for any pairing of donor and recipient populations, there is a maximum probability of identifying a match of a given quality for a random recipient, which cannot be exceeded even if registry size were infinite. This upper limit is a function of the frequency of blank (or unknown) alleles in the donor and recipient populations; the higher that frequency, the lower the maximum probability of achieving any given quality of match. The determinants of the probability of achieving a given quality of match with a given registry size are (1) the genetic heterogeneity within the recipient and donor populations, which increases the registry size required to achieve a given probability of matching, and (2) the degree of genetic homology between the donor and recipient populations, which increases the maximum probability of matching and also lowers registry size requirements. The method described here can be used to estimate donor pool size requirements using any donor and recipient populations for which HLA frequency data are available.

Automated critiquing of medical decision trees

The authors developed a decision tree-critiquing program (called BUNYAN) that identifies potential modeling errors in medical decision trees. The program's critiques are based on the structure of a decision problem, obtained from an abstract description specifying only the basic semantic categories of the model's components. A taxonomy of node and branch types supplies the primitive building blocks for representing decision trees. Bunyan detects potential problems in a model by matching general pattern expressions that refer to these primitives. A small set of general principles justifies critiquing rules that detect four categories of potential structural problems: impossible strategies, dominated strategies, unaccountable violations of symmetry, and omission of apparently reasonable strategies. Although critiquing based on structure alone has clear limitations, principled structural analysis constitutes the core of a methodology for reasoning about decision models.

Benign disappearance of ventricular thrombus after embolic stroke. A case report

The source for a patient's middle cerebral artery territory embolic stroke was found by echocardiography to be a left ventricular cardiac thrombus. The apical mass was large, pedunculated, and moved with systole into the ventricular cavity. The absence of ventricular dyskinesia was thought to favor a tumor, and surgery was considered before repeat echocardiography showed disappearance of the mass, making thrombus the likely diagnosis. No further embolic events occurred during or since the disappearance of the thrombus while on anticoagulation therapy. Serial echocardiography for change in or disappearance of a ventricular mass may be critical in distinguishing thrombus from tumor.

Inappropriate use of hospitals in a randomized trial of health insurance plans

We examined geographic variation in the rate of inappropriate hospitalization and the effect of cost sharing on that rate. The medical records of 1132 adults hospitalized in a randomized trial of health insurance plans were reviewed by two physicians who were blinded to the patients' insurance plan. They judged 23 percent of the admissions to be inappropriate and an additional 17 percent to have been avoidable by the use of ambulatory surgery. The percentage of inappropriate admissions varied among six sites (from 10 to 35 percent), but areas with low total admission rates did not necessarily have low proportions of inappropriate admissions. In plans with cost sharing for all services, 22 percent of admissions and 34 percent of hospital days were classified as inappropriate, as compared with 24 percent of admissions and 35 percent of hospital days in the plan under which care was free to the patient (these differences were not statistically significant). Our data show that a substantial fraction of hospitalization is potentially avoidable. Because cost sharing did not selectively reduce inappropriate hospitalization, it is important to develop other mechanisms to do so.

HLA-matched donor registries for bone marrow transplants. A decision analysis

Bone marrow transplantation has become an accepted treatment for severe aplastic anemia, acute myelogenous leukemia, and acute lymphocytic leukemia (20;35). Less commonly thalassemia (23), multiple myeloma (25), severe combined immune deficiencies (10), and chronic granulocytic leukemia (14) are also treated with marrow transplantation. Most donors are siblings whose HLA phenotype is identical to the patient's. Such donors, however, are available to only one third of patients (6;16;27;28;35). In the absence of an HLA-identical sibling, partially matched, haplo-identical, related donors (differing by only one major histocom-patibility antigen) can be found for another 7% of patients (1;6;27;37). A variety of lower quality matches also can be found but have not been used extensively. Advances in tissue typing and the treatment of graft versus host disease now allow physicians to consider matchedunrelatedindividuals as potential donors when a suitable related donor is not available (11;12;13;15;17;22).