The NUDGE trial pragmatic trial to enhance cardiovascular medication adherence: study protocol for a randomized controlled trial

Background

Nearly half of patients do not take their cardiovascular medications as prescribed, resulting in increased morbidity, mortality, and healthcare costs. Mobile and digital technologies for health promotion and disease self-management offer an opportunity to adapt behavioral “nudges” using ubiquitous mobile phone technology to facilitate medication adherence. The Nudge pragmatic clinical trial uses population-level pharmacy data to deliver nudges via mobile phone text messaging and an artificial intelligent interactive chat bot with the goal of improving medication adherence and patient outcomes in three integrated healthcare delivery systems.

Methods

The Theory of mHealth, the Expanded RE-AIM/PRISM, and the PRECIS-2 frameworks were used for program planning, implementation, and evaluation, along with a focus on dissemination and cost considerations. During the planning phase, the Nudge study team developed and piloted a technology-based nudge message and chat bot of optimized interactive content libraries for a range of diverse patients. Inclusion criteria are very broad and include patients in one of three diverse health systems who take medications to treat hypertension, atrial fibrillation, coronary artery disease, diabetes, or hyperlipidemia. A target of approximately 10,000 participants will be randomized to one of 4 study arms: usual care (no intervention), generic nudge (text reminder), optimized nudge, and optimized nudge plus interactive AI chat bot. The PRECIS-2 tool indicated that the study protocol is very pragmatic, although there is variability across PRECIS-2 dimensions.

Discussion

The primary effectiveness outcome is medication adherence defined by the proportion of days covered (PDC) using pharmacy refill data. Implementation outcomes are assessed using the RE-AIM framework, with a particular focus on reach, consistency of implementation, adaptations, cost, and maintenance/sustainability. The project has limitations including limited power to detect some subgroup effects, medication complications (bleeding), and longer-term outcomes (myocardial infarction). Strengths of the study include the diverse healthcare systems, a feasible and generalizable intervention, transparent reporting using established pragmatic research and implementation science frameworks, strong stakeholder engagement, and planning for dissemination and sustainment.

Trial registration

ClinicalTrials.govNCT03973931. Registered on 4 June 2019. The study was funded by the NIH; grant number is 4UH3HL144163-02 issued 4/5/19.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13063-021-05453-9.

Development of a measure of decision quality for implantable defibrillators

BACKGROUND

CMS reimbursement guidelines for implantable cardioverter-defibrillators (ICDs) include mandated shared decision making (SDM), but without any manner of assessing the quality of decisions made. We developed and tested a scale meant to assess patients' knowledge of and preferences specific to ICDs. Such a tool would assess these constructs in the clinical environment, targeting resources and support for patients considering a primary prevention ICD.

METHODS

Development of the ICD decision quality (ICD-DQ) scale included (1) item creation, (2) content validation using surveys of patients (n = 23) and clinicians (n = 31), and (3) examination of validity and reliability using a survey of patients who previously received an ICD (n = 295, response rate = 72%).

RESULTS

The final scale consists of 12 knowledge and 8 preference items. With respect to content validity, clinician and patient respondents agreed on the importance of 19 of 24 candidate knowledge items (79%), and 9 of 11 treatment preference items (81%). Knowledge items exhibited moderate internal validity (α = 0.62, 1 factor), strong test-retest reliability (mean % correct at first administration = 59%, 62% at follow-up, P > .1) and discriminant validity (59% correct for patients, 93% among cardiologists). Short versions of the ICD-DQ were developed for clinical settings, the scores from both of which correlated with the long version in this cohort (11-item (r = 0.90) and a 5-item (r = 0.75)).

CONCLUSIONS

The ICD-DQ fills a critical gap in measuring the quality of patients' ICD decisions. They may be used to evaluate the effectiveness of patient decision aids or the quality of SDM in clinical practice.

Trends in High- and Low-Value Cardiovascular Diagnostic Testing in Fee-for-Service Medicare, 2000-2016

IMPORTANCE

Owing to a rapid increase in rates of diagnostic cardiovascular testing in the 1990s and early 2000s, the Centers for Medicare & Medicaid Services implemented a series of payment changes intended to reduce overall spending on fee-for-service testing. Whether guideline-concordant testing has been subsequently affected is unknown to date.

OBJECTIVE

To determine whether changes in overall rates of use of diagnostic cardiovascular tests were associated with changes in high-value testing recommended by guidelines and low-value testing that is expected to provide minimal benefits.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study assessed a national 5% random sample of Medicare fee-for-service beneficiaries aged 65 to 95 years from January 1, 1999, through December 31, 2016. Data were analyzed from February 15, 2018, through August 15, 2019.

EXPOSURES

Eligibility to receive high-value testing (assessment of left ventricular systolic function among patients hospitalized with acute myocardial infarction or heart failure) and low-value testing (stress testing before low-risk noncardiac surgery and routine stress testing within 2 years of coronary revascularization not associated with acute care visits).

MAIN OUTCOMES AND MEASURES

Age- and sex-adjusted annual rates of overall, high-value, and low-value diagnostic cardiovascular testing.

RESULTS

Mean (SD) age was similar over time (75.57 [7.32] years in 2000-2003; 74.82 [7.79] years in 2012-2016); the proportion of women slightly declined over time (63.23% in 2000 to 2003; 57.27% in 2012 to 2016). The rate of overall diagnostic cardiovascular testing per 1000 patient-years among the 5% sample of Medicare beneficiaries increased from 275 in 2000 to 359 in 2008 (P < .001) and then declined to 316 in 2016 (P < .001). High-value testing increased steadily over the entire study period for patients with acute myocardial infarction (85.7% to 89.5%; P < .001) and heart failure (72.6% to 80.1%; P < .001). Low-value testing among patients undergoing low-risk surgery increased from 2.4% in 2000 to 3.8% in 2008 (P < .001) but then declined to 2.5% in 2016 (P < .001). Low-value testing within 2 years of coronary revascularization slightly increased from 47.4% in 2000 to 49.2% in 2003 (P = .03) but then declined to 30.8% in 2014 (P < .001).

CONCLUSIONS AND RELEVANCE

Rates of overall and low-value diagnostic cardiovascular testing appear to have declined considerably and rates of high-value testing have increased slightly. Payment changes intended to reduce spending on overall testing may not have adversely affected testing recommended by guidelines.

Association of Medical Liability Reform With Clinician Approach to Coronary Artery Disease Management

Importance

Physicians often report practicing defensive medicine to reduce malpractice risk, including performing expensive but marginally beneficial tests and procedures. Although there is little evidence that malpractice reform affects overall health care spending, it may influence physician behavior for specific conditions involving clinical uncertainty.

Objective

To examine whether reducing malpractice risk is associated with clinical decisions involving coronary artery disease testing and treatment.

Design, Setting, and Participants

Difference-in-differences design, comparing physician-specific changes in coronary artery disease testing and treatment in 9 new-cap states that adopted damage caps between 2003 and 2005 with 20 states without caps. We used the 5% national Medicare fee-for-service random sample between 1999 and 2013. Physicians (n = 75 801; 36 647 in new-cap states) who ordered or performed 2 or more coronary angiographies. Data were analyzed from June 2015 to January 2018.

Main Outcomes and Measures

Changes in ischemic evaluation rates for possible coronary artery disease, type of initial evaluation (stress testing or coronary angiography), progression from stress test to angiography, and progression from ischemic evaluation to revascularization (percutaneous coronary intervention or coronary artery bypass grafting).

Results

We studied 36 647 physicians in new-cap states and 39 154 physicians in no-cap states. New-cap states had younger populations, more minorities, lower per-capita incomes, fewer physicians per capita, and lower managed care penetration. Following cap adoption, new-cap physicians reduced invasive testing (angiography) as a first diagnostic test compared with control physicians (relative change, -24%; 95% CI, -40% to -7%; P = .005) with an offsetting increase in noninvasive stress testing (7.8%; 95% CI, -3.6% to 19.3%; P = .17), and referred fewer patients for angiography following stress testing (-21%; 95% CI, -40% to -2%; P = .03). New-cap physicians also reduced revascularization rates after ischemic evaluation (-23%; 95% CI, -40% to -4%; P = .02; driven by fewer percutaneous coronary interventions). Changes in overall ischemic evaluation rates were similar for new-cap and control physicians (-0.05%; 95% CI, -8.0% to 7.9%; P = .98).

Conclusions and Relevance

Physicians substantially altered their approach to coronary artery disease testing and follow-up after initial ischemic evaluations following adoption of damage caps. They performed a similar number of ischemic evaluations but conducted fewer initial left heart catheterizations, referred fewer stress-tested patients for left heart catheterizations, and referred fewer patients for revascularization. These findings suggest that physicians tolerate greater clinical uncertainty in coronary artery disease testing and treatment if they face lower malpractice risk.

Coronary artery disease severity modifies associations between glycemic control and both mortality and myocardial infarction

AIMS

This study examined whether the association between hemoglobin A1c (HbA1c) and short-term clinical outcomes is moderated by CAD severity.

METHODS

We studied 17,394 US Veterans with type 2 diabetes who underwent elective cardiac catheterization between 2005 and 2013. CAD severity was categorized as obstructive, non-obstructive, or no CAD. Using multivariable Cox proportional hazards regression, we assessed associations between time-varying HbA1c and two-year all-cause mortality and non-fatal MI, with an interaction term between HbA1c and CAD severity.

RESULTS

61%, 22%, and 17% of participants had obstructive, non-obstructive, and no CAD, respectively. CAD severity modified the relationship between HbA1c and each outcome (interaction p-value 0.0005 for mortality and <0.0001 for MI). Low HbA1c (<42 mmol/mol) was associated with increased mortality, relative to HbA1c of 48-52 mmol/mol, in individuals with obstructive CAD (HR 1.52 [1.17, 1.97]) and non-obstructive CAD (HR 2.61 [1.61, 4.23]), but not in those with no CAD (HR 0.91 [0.46, 1.79]). In contrast, higher HbA1c levels (≥53 mmol/mol) were associated with increased MI risk only in individuals with obstructive CAD.

CONCLUSIONS

The associations between HbA1c and mortality and MI were moderated by CAD severity. Measures of cardiovascular disease severity may inform optimal individualized diabetes management.

Abstract 28: Variation in High-Value Cardiovascular Diagnostic Testing: Patient, Payer, and Hospital Effects

Background: In patients hospitalized with 1) newly diagnosed heart failure (HF) or 2) acute myocardial infarction (AMI), assessment of left ventricular systolic function is a high-value test supported by guidelines. We examined the degree to which patient-, payer-, and hospital-level characteristics impact use of testing.

Methods: We analyzed data from the Colorado All-Payer Claims Database, a repository of billing claims from all insurers who provide care in the state. We identified all patients with an index hospitalization for HF and AMI from 2010 to 2014. We excluded patients with a prior diagnosis of HF, and hospitals with fewer than 40 HF or AMI hospitalizations. We determined whether patients had a systolic function assessment performed within 60 days of hospitalization. We calculated adjusted rates of testing at the hospital level, and assessed for correlation of rates between HF and AMI patients. We used multilevel logistic regression to assess patient- and payer- characteristics associated with testing, and used median odds ratios to determine the residual variation in testing attributable to hospitals.

Results: We identified 9,516 patients with HF and 10,315 patients with AMI (mean age 73 years, 48% women) among 36 hospitals. Overall, 74% of HF patients and 73% of AMI patients received testing. Testing rates among hospitals ranged from 56% to 82% for HF and from 42% to 83% for AMI (Figure). Correlation of testing rates for AMI and HF patients among hospitals was moderate (Spearman r=0.58; p<.001). Medicaid insurance was associated with lower likelihood of testing for both AMI and HF (ORs 0.77 [0.67-0.88] and 0.54 [0.47-0.62]; both p<.001). After multivariable adjustment, use of testing across sites varied by a median odds ratio of 1.39 [1.28-1.49] for AMI patients and of 1.25 [1.17-1.34] for HF patients, meaning that on average, patients had 1.39 and 1.25 higher odds of being tested if they received care at a higher performing hospital.

Conclusions: Despite adjustment for patient- and payer-level characteristics, there is 1) significant residual variation in use of high-value cardiac testing and 2) correlation in testing rates for AMI and HF patients among hospitals. These results suggest that hospital-level characteristics and care processes may have a strong influence on use of high-value testing.

Postdischarge Bleeding After Percutaneous Coronary Intervention and Subsequent Mortality and Myocardial Infarction: Insights From the HMO Research Network-Stent Registry

BACKGROUND

Bleeding after hospital discharge from percutaneous coronary intervention (PCI) is associated with increased risk of subsequent myocardial infarction (MI) and death; however, the timing of adverse events after these bleeding events is poorly understood. Defining this relationship may help clinicians identify critical periods when patients are at highest risk.

METHODS AND RESULTS

All patients undergoing PCI from 2004 to 2007 who survived to hospital discharge without a bleeding event were identified from the HMO Research Network-Stent (HMORN-Stent) Registry. Postdischarge rates and timing of bleeding-related hospitalizations, MI, and death were defined. We then assessed the association between postdischarge bleeding-related hospitalizations with death and MI using Cox proportional hazards models. Among 8137 post-PCI patients surviving to hospital discharge without in-hospital bleeding, 391 (4.8%) had bleeding-related hospitalization after discharge, with the highest incidence of bleeding-related hospitalizations occurring within 30 days of discharge (n=79, 20.2%). Postdischarge bleeding-related hospitalization after PCI was associated with subsequent death or MI (hazard ratio, 3.09; 95% confidence interval, 2.41-3.96), with the highest risk for death or MI occurring in the first 60 days after bleeding-related hospitalization (hazard ratio, 7.16; confidence interval, 3.93-13.05).

CONCLUSIONS

Approximately 1 in 20 post-PCI patients are readmitted for bleeding, with the highest incidence occurring within 30 days of discharge. Patients having postdischarge bleeding are at increased risk for subsequent death or MI, with the highest risk occurring within the first 60 days after a bleeding-related hospitalization. These findings suggest a critical period after bleeding events when patients are most vulnerable for further adverse events.

Evaluation of an Electronic Critical Drug Interaction Program Coupled with Active Pharmacist Intervention

Background:

Failure to detect significant drug interactions may result in adverse outcomes. While proper screening and management of drug interactions can prevent the majority of adverse events, studies indicate that current practice is suboptimal. In the last quarter of 2001, physicians and pharmacists in Kaiser Permanente Colorado, a group model health maintenance organization, developed an electronic critical drug interaction alert program (CDIX). Electronic screening was coupled with active intervention to prevent dispensing of critically interacting drug combinations.

Objective:

To assess the impact of CDIX on the co-dispensing ol critically interacting drug combinations.

Methods:

A physician and team of outpatient pharmacists and clinical pharmacy staff developed a condensed list of critical drug interactions (8 drug combinations) to be included in the evaluation of CDIX. Monthly electronic outpatient pharmacy data were collected 20 months before and 37 months after CDIX implementation, with no lag period following implementation. Univariate analyses were completed to compare baseline subject characteristics of the pre- and post-CDIX groups using χ2 and Wilcoxon Rank Sum tests. Interrupted time series analysis was used to estimate changes in the rates of critical drug interactions.

Results:

Three hundred sixty-seven instances of co-dispensing were observed in 348 subjects during the pre-CDIX period and 256 instances of co-dispensing were observed in 248 subjects during the post-CDIX period. Following CDIX implementation, the overall rate of co-dispensing dropped abruptly from 21.3 to 14.7 per 10,000 prescriptions, representing a relative decrease in co-dispensing of 31% from the month before CDIX implementation (p = 0.0125). Significant reductions in co-dispensing were noted for 7 of the 8 drug class combinations.

Conclusions:

Employing an intervention system that limits electronic alerts regarding drug interactions to those deemed critical but that also requires pharmacist intervention and collaboration with the prescriber decreases the number of critical drug interactions dispensed.

Abstract 136: Recidivism to Uncontrolled Blood Pressure in Patients with Previously Controlled Hypertension

Objectives: Control of hypertension has improved nationally with focus on identifying and treating elevated blood pressures (BP) to guideline recommended levels. However, once BP control is achieved, the frequency in which BP falls out of control and the factors associated with BP recidivism is unknown. The objective of this study is to examine rates and predictors of blood pressure recidivism in adults with controlled hypertension.

Study Design: A retrospective cohort study in two large, integrated health care systems in Minnesota and Colorado.

Methods: Patients with a prior diagnosis of hypertension based on a combination of ICD-9 codes, receipt of anti-hypertensive medications and/or elevated blood pressure readings were eligible to be included. We defined controlled hypertension as normotensive blood pressure (BP) readings (less than 140/90 mmHg or less than 130/80 mmHg if coexistent diabetes or chronic kidney disease present) at 2 consecutive primary care visits. Following these visits among patients with controlled BP, we followed patients for BP recidivism defined by the mean of the last 2 blood pressure readings greater than 140/90 mm Hg or 130/80 mm Hg for those with diabetes or chronic kidney disease during a mean follow-up period of 13.2 months .

Results: A total of 22,275 patients with controlled hypertension were included in this study. The proportion of patients with hypertension recidivism was 16.4%. A linear increase in blood pressure recidivism was noted with time between the index visit and last observed blood pressure reading. Major predictors of recidivism included female gender (OR 1.14, p = 0.02), history of diabetes (OR 2.32, p < 0.001) and black race (OR 1.37, p = 0.02). Age 50-64 displayed a protective effect against recidivism (OR 0.80, p = 0.02).

Conclusions: Hypertensive recidivism occurs in a significant portion of patients with controlled hypertension. Patient factors associated with recidivism include age, race and prior history of diabetes among other characteristics. Strategies to minimize hypertension recidivism have significant potential to improve overall levels of blood pressure control and hypertension related quality measures from a health care systems perspective.

Abstract 24: Non-Obstructive Coronary Artery Disease Is Not Benign: Insights from the VA CART Program on the Association between Non-Obstructive Disease and Cardiac Events

Introduction: The traditional focus of cardiac care on obstructive (>70% stenosis) CAD potentially distracts from the risks inherent in non-obstructive CAD. However, surprisingly little is known about non-obstructive CAD outcomes. Therefore, we determined the association between non-obstructive CAD and cardiovascular outcomes.

Methods: Using the national VA CART program, we studied all veterans undergoing elective coronary angiography for angina between October 2007 and September 2012. Patients were categorized by CAD extent (none (no stenosis >20%), non-obstructive (no stenosis >=70%), obstructive (any stenosis >=70%)) and distribution (1, 2, or 3 vessel), and assessed for major adverse cardiac events (MACE), defined as all-cause mortality and MI. We adjusted for demographic, clinical, and treatment factors using Cox proportional hazards modeling. Secondary analyses sub-divided non-obstructive CAD into mild (20-49% stenosis) and moderate (50-69% stenosis) disease.

Results: During the study period, 40,872 veterans underwent catheterization. Of these, 8411 (20.6%) had no CAD, 5219 (17.7%) had 1V non-obstructive CAD, 3034 (10.3%) had 2V non-obstructive CAD, 1388 (4.7%) had 3V non-obstructive CAD, 8588 (29.1%) had 1V obstructive, 5227 (17.7%) had 2V obstructive, and 6017 (20.4%) had 3V/LM obstructive CAD. MACE rates progressively increased with increasing CAD severity (Figure). This association persisted after risk adjustment (HR 1.28 (1.08, 1.51) for 1V non-obstructive, 1.29 (1.08, 1.52) 2V non-obstructive, 1.44 (1.12, 1.86) 3V non-obstructive, 1.93 (1.64, 2.28) 1V obstructive, 2.73 (2.28, 3.27) 2V obstructive, and 2.98 (2.52, 3.53) 3V/LM obstructive CAD)). A trend toward higher MACE in moderate 3V non-obstructive compared to 1V obstructive CAD (HR 1.34 (0.71, 2.52)) was noted.

Conclusions: Non-obstructive CAD, relative to no CAD, is associated with 28-44% higher odds of MACE. MACE risk progressively increases by CAD extent, rather than abruptly increasing between non-obstructive and obstructive CAD. The risks of adverse events were similar for 3V non-obstructive CAD and 1V obstructive CAD, highlighting the limitations of a dichotomous characterization of angiographic CAD and a need to recognize the risks inherent in non-obstructive CAD.

Estimating the effects of time-varying exposures in observational studies using Cox models with stabilized weights adjustment

PURPOSE

Assessing the safety and effectiveness of medical products with observational electronic medical record data is challenging when the treatment is time-varying. The objective of this paper is to develop a Cox model stratified by event times with stabilized weights (SWs) adjustment to examine the effect of time-varying treatment in observational studies.

METHODS

Time-varying SWs are calculated at unique event times and are used in a Cox model stratified by event times to estimate the effect of time-varying treatment. We applied this method in examining the effect of an antiplatelet agent, clopidogrel, on events, including bleeding, myocardial infarction, and death after a drug-eluting stent was implanted in coronary artery. Clopidogrel use may change over time on the basis of patients' behavior (e.g., non-adherence) and physicians' recommendations (e.g., end of duration of therapy). We also compared the results with those from a Cox model for counting processes adjusting for all covariates used in creating SWs.

RESULTS

We demonstrate that the (i) results from the stratified Cox model without SWs adjustment and the Cox model for counting processes without covariate adjustment are identical in analyzing the clopidogrel data; and (ii) the effects of clopidogrel on bleeding, myocardial infarction, and death are larger in the stratified Cox model with SWs adjustment compared with those from the Cox model for counting processes with covariate adjustment.

CONCLUSIONS

The Cox model stratified by event times with time-varying SWs adjustment is useful in estimating the effect of time-varying treatments in observational studies while balancing for known confounders.

Prehypertension and hypertension in community-based pediatric practice

OBJECTIVE

To examine the prevalence of prehypertension and hypertension among children receiving well-child care in community-based practices.

METHODS

Children aged 3 to 17 years with measurements of height, weight, and blood pressure (BP) obtained at an initial (index) well-child visit between July 2007 and December 2009 were included in this retrospective cohort study across 3 large, integrated health care delivery systems. Index BP classification was based on the Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents: normal BP, <90th percentile; prehypertension, 90th to 94th percentile; hypertension, 3 BP measurements ≥95th percentile (index and 2 subsequent consecutive visits).

RESULTS

The cohort included 199 513 children (24.3% aged 3-5 years, 34.5% aged 6-11 years, and 41.2% aged 12-17 years) with substantial racial/ethnic diversity (35.9% white, 7.8% black, 17.6% Hispanic, 11.7% Asian/Pacific Islander, and 27.0% other/unknown race). At the index visit, 81.9% of participants were normotensive, 12.7% had prehypertension, and 5.4% had a BP in the hypertension range (≥95th percentile). Of the 10 848 children with an index hypertensive BP level, 3.8% of those with a follow-up BP measurement had confirmed hypertension (estimated 0.3% prevalence). Increasing age and BMI were significantly associated with prehypertension and confirmed hypertension (P < .001 for trend). Among racial/ethnic groups, blacks and Asians had the highest prevalence of hypertension.

CONCLUSIONS

The prevalence of hypertension in this community-based study is lower than previously reported from school-based studies. With the size and diversity of this cohort, these results suggest the prevalence of hypertension in children may actually be lower than previously reported.

Health system factors and antihypertensive adherence in a racially and ethnically diverse cohort of new users

BACKGROUND

The purpose of this study was to identify potential health system solutions to suboptimal use of antihypertensive therapy in a diverse cohort of patients initiating treatment.

METHODS

Using a hypertension registry at Kaiser Permanente Northern California, we conducted a retrospective cohort study of 44 167 adults (age, ≥18 years) with hypertension who were new users of antihypertensive therapy in 2008. We used multivariate logistic regression analysis to model the relationships between race/ethnicity, specific health system factors, and early nonpersistence (failing to refill the first prescription within 90 days) and nonadherence (<80% of days covered during the 12 months following the start of treatment), respectively, controlling for sociodemographic and clinical risk factors.

RESULTS

More than 30% of patients were early nonpersistent and 1 in 5 were nonadherent to therapy. Nonwhites were more likely to exhibit both types of suboptimal medication-taking behavior compared with whites. In logistic regression models adjusted for sociodemographic, clinical, and health system factors, nonwhite race was associated with early nonpersistence (black: odds ratio, 1.56 [95% CI, 1.43-1.70]; Asian: 1.40 [1.29-1.51]; Hispanic: 1.46 [1.35-1.57]) and nonadherence (black: 1.55 [1.37-1.77]; Asian: 1.13 [1.00-1.28]; Hispanic: 1.46 [1.31-1.63]). The likelihood of early nonpersistence varied between Asians and Hispanics by choice of first-line therapy. In addition, racial and ethnic differences in nonadherence were appreciably attenuated when medication co-payment and mail-order pharmacy use were accounted for in the models.

CONCLUSIONS

Racial/ethnic differences in medication-taking behavior occur early in the course of treatment. However, health system strategies designed to reduce patient co-payments, ease access to medications, and optimize the choice of initial therapy may be effective tools in narrowing persistent gaps in the use of these and other clinically effective therapies.

Extension of Kaplan-Meier methods in observational studies with time-varying treatment

OBJECTIVES

Inverse probability of treatment weighted Kaplan-Meier estimates have been developed to compare two treatments in the presence of confounders in observational studies. Recently, stabilized weights were developed to reduce the influence of extreme inverse probability of treatment-weighted weights in estimating treatment effects. The objective of this research was to use adjusted Kaplan-Meier estimates and modified log-rank and Wilcoxon tests to examine the effect of a treatment that varies over time in an observational study.

METHODS

We proposed stabilized weight adjusted Kaplan-Meier estimates and modified log-rank and Wilcoxon tests when the treatment was time-varying over the follow-up period. We applied these new methods in examining the effect of an anti-platelet agent, clopidogrel, on subsequent events, including bleeding, myocardial infarction, and death after a drug-eluting stent was implanted into a coronary artery. In this population, clopidogrel use may change over time based on a patient's behavior (e.g., nonadherence) and physicians' recommendations (e.g., end of duration of therapy). Consequently, clopidogrel use was treated as a time-varying variable.

RESULTS

We demonstrate that 1) the sample sizes at three chosen time points are almost identical in the original and weighted datasets; and 2) the covariates between patients on and off clopidogrel were well balanced after stabilized weights were applied to the original samples.

CONCLUSIONS

The stabilized weight-adjusted Kaplan-Meier estimates and modified log-rank and Wilcoxon tests are useful in presenting and comparing survival functions for time-varying treatments in observational studies while adjusting for known confounders.

Abstract 3: Treatment Intensification and Medication Adherence Are Factors in Improving Blood Pressure Trajectories Among Newly Diagnosed CAD Patients

Background: Blood pressure (BP) control among coronary artery disease (CAD) patients is sub-optimal in clinical practice and may be due to gaps in treatment intensification and medication adherence. However, longitudinal studies evaluating these relationships are limited.

Methods: We assessed BP trajectories among managed care patients with hypertension and incident CAD. Patients were stratified by target BP goal (<140/90 or <130/80 among those with diabetes or renal disease) and BP trajectories were modeled over the year following CAD diagnosis. Treatment intensification (increase in BP therapies in the setting of an elevated BP) and medication adherence (percentage of days covered with BP therapies) were evaluated in multivariable models adjusting for patient variables.

Results: 2763 patients had a <140/90 BP goal and 9890 had a <130/80 BP goal. Within each BP goal group, four trajectories were identified - persistently uncontrolled (Group 1), uncontrolled to controlled (Group 2), controlled to uncontrolled (Group 3), and persistently controlled (Group 4). In multivariable models comparing Group 2 to Group 1, treatment intensification was associated with improved BP control over the year for both groups (OR for <140/90 BP goal 0.75, 95% CI 0.66, 0.84; OR for <130/80 BP goal 0.93, 95% CI 0.89, 0.98). Medication adherence was associated with improved BP control among the 130/80 target BP group only (OR for <140/90 BP goal 0.79, 95% CI 0.38, 1.69; OR for <130/80 BP goal 0.66, 95% CI 0.49, 0.9).

Conclusions: In this managed care patient population, treatment intensification and medication adherence significantly affect BP trajectories in the year following CAD diagnosis.

Patients' perspectives of a multifaceted intervention with a focus on technology: a qualitative analysis

BACKGROUND

Few studies have investigated the effectiveness of multifaceted interventions from the study participants' perspective. We conducted qualitative interviews to understand patients' experiences with a multifaceted blood pressure (BP) control intervention involving interactive voice response technology, home BP monitoring, and pharmacist-led BP management. In the randomized study, the intervention resulted in clinically significant decreases in BP.

METHODS AND RESULTS

We used insights generated from in-depth interviews from all study participants randomly assigned to the multifaceted intervention or usual care (n=146) to create a model explaining the observed improvements in health behavior and clinical outcomes. The data were analyzed using qualitative content analysis methods and consultative and reflexive team analysis. Six explanatory factors emerged from the patients' interviews: (1) improved relationships with medical personnel; (2) increased knowledge of hypertension; (3) increased participation in their health care and personal empowerment; (4) greater understanding of the impact of health behavior on BP; (5) high satisfaction with technology used in the intervention; and, for some patients, (6) increased health care utilization. Eighty-six percent of the intervention patients and 62% of the usual care patients stated that study participation had a positive effect on them. Of those expressing a positive effect, 68% (intervention) and 55% (usual care) reached their systolic BP goal.

CONCLUSIONS

Establishing bidirectional conversations between patients and providers is a key element of successful hypertension management. Home BP monitoring coupled with interactive voice response technology reporting facilitates such conversations.

Impaired heart rate recovery is associated with new-onset atrial fibrillation: a prospective cohort study

Background

Autonomic dysfunction appears to play a significant role in the development of atrial fibrillation (AF), and impaired heart rate recovery (HRR) during exercise treadmill testing (ETT) is a known marker for autonomic dysfunction. However, whether impaired HRR is associated with incident AF is unknown. We studied the association of impaired HRR with the development of incident AF, after controlling for demographic and clinical confounders.

Methods

We studied 8236 patients referred for ETT between 2001 and 2004, and without a prior history of AF. Patients were categorized by normal or impaired HRR on ETT. The primary outcome was the development of AF. Cox proportional hazards modeling was used to control for demographic and clinical characteristics. Secondary analyses exploring a continuous relationship between impaired HRR and AF, and exploring interactions between cardiac medication use, HRR, and AF were also conducted.

Results

After adjustment, patients with impaired HRR were more likely to develop AF than patients with normal HRR (HR 1.43, 95% confidence interval (CI) 1.06, 1.93). In addition, there was a linear trend between impaired HRR and AF (HR 1.05 for each decreasing BPM in HRR, 95% CI 0.99, 1.11). No interactions between cardiac medications, HRR, and AF were noted.

Conclusion

Patients with impaired HRR on ETT were more likely to develop new-onset AF, as compared to patients with normal HRR. These findings support the hypothesis that autonomic dysfunction mediates the development of AF, and suggest that interventions known to improve HRR, such as exercise training, may delay or prevent AF.

Factors related to adherence to statin therapy

BACKGROUND

Retrospective database analyses have revealed that 50% of patients receiving statins discontinue therapy after one year of treatment. Typically, these data do not focus on patient-specific reasons for discontinuation.

OBJECTIVE

To examine the reasons that patients discontinue statins and compare the patient and clinical factors of those who do and do not discontinue therapy.

METHODS

All patients with a new statin prescription between January 1, 2004, and March 31, 2004, were identified through pharmacy claims. Patients who had discontinued and continued statin therapy were identified. Medical records were reviewed to determine whether there were documented reasons for statin discontinuation. Subsequently, telephone surveys addressing statin knowledge, relationships, communication with healthcare providers, and general health status were conducted.

RESULTS

At one year, 47.5% (n = 671) of patients had obtained fewer than 80% of the refills of their prescribed statin. We reviewed 435 medical records and conducted 255 patient surveys. A total of 29.9% of discontinuers had reasons documented in the medical record. Compared with continuers, fewer discontinuers had follow-up and/or laboratory visits with a provider within 6 months after the start of statin therapy. The surveys indicated that more continuers than discontinuers trusted their providers (p < 0.05) and felt that providers had adequate knowledge to answer their questions (p < 0.001). In contrast, more discontinuers felt the statin was of limited benefit/unsure of the benefit (p < 0.001) and believed that their providers were not interested in their input on their medical condition (p < 0.01).

CONCLUSIONS

Utilizing pharmacy claims records alone to determine statin nonadherence may not only overestimate the percentage of patients who are nonadherent, but also prevent healthcare providers from understanding the reasons that patients discontinue or continue statin therapy. Statin adherence is complex and affected by several factors. Interventions to improve adherence should focus on patient communications, education, and follow-up.

Association between parental depression and children's health care use

OBJECTIVE

The objective of this study was to determine the association between parental depression and pediatric health care use patterns.

METHODS

We selected all children who were 0 to 17 years of age, enrolled in Kaiser Permanente of Colorado during the study period July 1997 to December 2002, and linked to at least 1 parent/subscriber who was enrolled for at least 6 months during that period. Unexposed children were selected from a pool of children whose parents did not have a depression diagnosis. Outcome measures were derived from the child's payment files and electronic medical charts and included 5 categories of use: well-child-care visits, sick visits to primary care departments, specialty clinic visits, emergency department visits, and inpatient visits. We compared the rate of use per enrollment month for these 5 categories between exposed and unexposed children within each of the 5 age strata.

RESULTS

Our study population had 24,391 exposed and 45,274 age-matched, unexposed children. For the outcome of well-child-care visits, teenagers showed decreased rates of visits among exposed children. The rate of specialty department visits was higher in exposed children in the 4 oldest age groups. The rates of both emergency department visits and sick visits to primary care departments were higher for exposed children across all 5 age categories. The rate of inpatient visits was higher among exposed children in 2 of the 5 age groups.

CONCLUSIONS

Overall, having at least 1 depressed parent is associated with greater rate of emergency department and sick visits across all age groups, greater use of inpatient and specialty services in some age groups, and a lower rate of well-child-care visits among 13- to 17-year-olds. This pattern of increased use of expensive resources and decreased use of preventive services represents one of the hidden costs of adult depression.

How safe is triage by an after-hours telephone call center?

OBJECTIVES

Our goals were to assess (1) compliance with nurse disposition recommendations, (2) frequency of death or potential underreferral associated with hospitalization within 24 hours after a call, and (3) factors associated with potential underreferral, for children receiving care within an integrated health care delivery organization who were triaged by a pediatric after-hours call center.

METHODS

The study population included all pediatric patients enrolled in Kaiser Permanente Colorado whose families called the Children's Hospital after-hours call center in Denver, Colorado, during the period between October 1, 1999, and March 31, 2003. Postcall disposition recommendations were categorized as urgent (visit within 4 hours), next day (visit in > 4 hours but within 24 hours), later visit (visit in > 24 hours), or home care (care at home without a visit). Compliance with the nurses' triage disposition recommendations was calculated as the proportion of cases for which utilization data matched the disposition recommendations.

RESULTS

Of the 32,968 eligible calls during the study period, 21% received urgent, 27% next day, 4% later visit, and 48% home care disposition recommendations. Rates of compliance with both urgent and home care disposition recommendations were 74%, and the rate of compliance with next day recommendations was 44%. No deaths occurred within < 1 week after the after-hours calls. The rate of potential underreferral with subsequent hospitalization was 0.2%, or 1 case per 599 triaged calls. In multivariate modeling, age of < 6 weeks or > 12 years and being triaged after 11 pm were associated with higher rates of potential under-referral.

CONCLUSIONS

Approximately three fourths of families complied with recommendations for their child to be evaluated urgently or to be treated at home, with much lower rates of compliance with intermediate dispositions. The rate of potential underreferral with hospitalization was low, and age and time of call triage were associated with this outcome.

Telephone screening for hazardous drinking among injured patients seen in acute care clinics: feasibility study

AIMS

We evaluated the effectiveness of telephoning injured patients after discharge, compared with contacting them in the clinic during the acute care visit, for screening for hazardous drinking and eliciting willingness to participate in a lifestyle intervention trial.

METHODS

We conducted a quasi-randomized controlled trial among acutely injured adult patients in trauma and acute care clinics, assigning telephone and clinic screening strategies systematically by week. During telephone weeks, we mailed study information to patients identified from computerized records, then telephoned them. During clinic weeks, researchers recruited patients awaiting care. We screened for hazardous drinking using the AUDIT-C (Alcohol Use Disorders Identification Test-C). We examined the proportion of all injured adult patients who were screened, the proportion of screened patients with hazardous drinking (AUDIT-C score >or=4), and the proportion willing to participate in a (hypothetical) lifestyle intervention trial. Differences were analysed with non-linear mixed models using generalized estimating equations, controlling for age, sex, and facility. Levers and barriers to screening were explored through structured interviews with research staff.

RESULTS

We enrolled 29% (469/1,609) of all injured adult patients and 76% of injured patients contacted and found to be eligible. Of screened patients, 23.1% screened positive for hazardous drinking. Telephone and clinic contact were equally effective for screening patients (OR = 1.05; 95% CI = 0.59-1.87), identifying hazardous drinking (OR=0.97; 95% CI = 0.54-1.74), and eliciting willingness to participate in an intervention trial (OR=1.49; 95% CI = 0.97-2.30). Clinic site modified results: telephone was more effective than clinic contact for screening urban patients (OR=1.99; 95% CI = 1.36-2.93), but less effective for screening suburban patients (OR = 0.70; 95% CI = 0.69-0.71). Barriers to clinic screening included lack of clinic staff support, time constraints, and difficulty recruiting elderly or acutely distressed patients. Barriers to telephone screening included erroneous contact information and failure to answer the telephone.

CONCLUSIONS

Telephone screening is a feasible and efficient method for screening moderately injured adult patients for hazardous drinking, but characteristics of the clinical site (including personnel) influence its effectiveness. Trauma and acute care clinics are likely to be fruitful sites for identification of patients with hazardous drinking, whether for enrollment into brief intervention trials or treatment programmes.

Evaluation of Clinician-Reported Adherence to Centers for Disease Control and Prevention Guidelines for the Treatment of Chlamydia trachomatis in Two U.S. Health Plans

OBJECTIVE

The objective of this study was to assess clinician adherence to Centers for Disease Control and Prevention-recommended treatments for Chlamydia trachomatis (CT) in two health plans.

STUDY DESIGN

Using hypothetical scenarios, a 1999-2000 mail survey questioned clinicians about how they would treat a cervicitis patient (CT and gonorrhea treatment recommended) and two patients with laboratory-confirmed CT: an injection drug user (single-dose azithromycin promotes adherence) and a pregnant patient (nonteratogenic drugs recommended).

RESULTS

Seven hundred forty-three (82%) of the 907 nonretired clinicians receiving the survey completed it. Eighty-one percent (N=599) reported providing recent CT care. Of these, 70.1% reported they would presumptively treat patients with cervicitis for CT and gonorrhea, 17.1% for CT only, and 11.7% for neither pathogen. Of the 580 clinicians addressing drug injectors, 61.7% reported they would prescribe azithromycin. Most (88.8%) of the 343 clinicians seeing pregnant patients reported they would prescribe Centers for Disease Control and Prevention (CDC)-recommended antibiotics. Reported adherence varied by clinician specialty and sources of treatment guidance.

CONCLUSIONS

Most clinicians reported treatment consistent with CDC guidelines.

Hospital improvement in time to reperfusion in patients with acute myocardial infarction, 1999 to 2002

OBJECTIVES

The purpose of this study was to analyze recent trends in door-to-reperfusion time and to identify hospital characteristics associated with improved performance.

BACKGROUND

Rapid reperfusion improves survival for patients with acute ST-segment elevation myocardial infarction (STEMI).

METHODS

In this retrospective observational study from the National Registry of Myocardial Infarction (NRMI)-3 and -4, between 1999 and 2002, we analyzed door-to-needle and door-to-balloon times in patients admitted with STEMI and receiving fibrinolytic therapy (n = 68,439 patients in 1,015 hospitals) or percutaneous coronary intervention (n = 33,647 patients in 421 hospitals) within 6 h of hospital arrival.

RESULTS

In 1999, only 46% of the patients in the fibrinolytic therapy cohort were treated within the recommended 30-min door-to-needle time; only 35% of the patients in the percutaneous coronary intervention cohort were treated within the recommended 90-min door-to-balloon time. Improvement in these times to reperfusion over the four-year study period was not statistically significant (door-to-needle: -0.01 min/year, 95% confidence interval [CI] -0.24 to +0.23, p > 0.9; door-to-balloon: -0.57 min/year, 95% CI -1.24 to +0.10, p = 0.09). Only 33% (337 of 1,015) of hospitals improved door-to-needle time by more than one min/year, and 26% (110 of 421) improved door-to-balloon time by more than three min/year. No hospital characteristic was significantly associated with improvement in door-to-needle time. Only high annual percutaneous coronary intervention volume and location in New England were significantly associated with greater improvement in door-to-balloon time.

CONCLUSIONS

Fewer than one-half of patients with STEMI receive reperfusion in the recommended door-to-needle or door-to-balloon time, and mean time to reperfusion has not decreased significantly in recent years. Relatively few hospitals have shown substantial improvement.

Use of emergency medical services in acute myocardial infarction and subsequent quality of care: observations from the National Registry of Myocardial Infarction 2

BACKGROUND

National practice guidelines strongly recommend activation of the 9-1-1 Emergency Medical Systems (EMS) by patients with symptoms consistent with an acute myocardial infarction (MI). We examined use of the EMS in the United States and ascertained the factors that may influence its use by patients with acute MI.

METHODS AND RESULTS

From June 1994 to March 1998, the National Registry of Myocardial Infarction 2 enrolled 772 586 patients hospitalized with MI. We excluded those who transferred in, arrived at the hospital >6 hours from symptom onset, or who were in cardiogenic shock. We compared baseline characteristics and initial management for patients who arrived by ambulance versus self-transport. EMS was used in 53.4% of patients with MI, a proportion that did not vary significantly over the 4-year study period. Nonusers of the EMS were on average younger, male, and at relatively lower risk on presentation. In addition, payer status was significantly associated with EMS use. Use of EMS was independently associated with slightly wider use of acute reperfusion therapies and faster time intervals from door to fibrinolytic therapy (12.1 minutes faster, P<0.001) or to urgent PTCA (31.2 minutes faster, P<0.001).

CONCLUSIONS

Only half of patients with MI were transported to the hospital by ambulance, and these patients had greater and significantly faster receipt of initial reperfusion therapies. Wider use of EMS by patients with suspected MI may offer considerable opportunity for improvement in public health.

The hip in Stickler syndrome

Stickler syndrome is an autosomal dominant connective tissue disorder with a prevalence similar to that of Marfan syndrome. No previous study has examined hip pain or abnormalities in a large series of patients with Stickler syndrome. The purpose of this study was to describe hip abnormalities and their correlation with age and chronic hip pain in a cohort of 51 patients followed at the National Institutes of Health. Ten percent of patients had protrusio acetabuli, 21% coxa valga, and 34% of adults had hip osteoarthritis. Sixty-three percent of all patients and 79% of adults had chronic hip pain. In addition, 16% of adult patients had a history of femoral head failure during youth. Arthritic changes and adult age were associated with hip pain. In summary, hip abnormalities are commonly observed in Stickler syndrome. Young patients require careful evaluation of hip pain, and regular screening of children with Stickler syndrome may be indicated for early detection of hip complications.

Interpretation of Emergency Department radiographs: a comparison of emergency medicine physicians with radiologists, residents with faculty, and film with digital display

OBJECTIVE

We determined the relative value of teleradiology and radiology resident coverage of the emergency department by measuring and comparing the effects of physician specialty, training level, and image display method on accuracy of radiograph interpretation.

MATERIALS AND METHODS

A sample of four faculty emergency medicine physicians, four emergency medicine residents, four faculty radiologists, and four radiology residents participated in our study. Each physician interpreted 120 radiographs, approximately half containing a clinically important index finding. Radiographs were interpreted using the original films and high-resolution digital monitors. Accuracy of radiograph interpretation was measured as the area under the physicians' receiver operating characteristic (ROC) curves.

RESULTS

The area under the ROC curve was 0.15 (95% confidence interval [CI], 0.10-0.20) greater for radiologists than for emergency medicine physicians, 0.07 (95% CI, 0.02-0.12) greater for faculty than for residents, and 0.07 (95% CI, 0.02-0.12) greater for films than for video monitors. Using these results, we estimated that teleradiology coverage by faculty radiologists would add 0.09 (95% CI, 0.03-0.15) to the area under the ROC curve for radiograph interpretation by emergency medicine faculty alone, and radiology resident coverage would add 0.08 (95% CI, 0.02-0.14) to this area.

CONCLUSION

We observed significant differences between the interpretation of radiographs on film and on digital monitors. However, we observed differences of equal or greater magnitude associated with the training level and physician specialty of each observer. In evaluating teleradiology services, observer characteristics must be considered in addition to the quality of image display.

Universal versus targeted screening of infants for sickle cell disease: a cost-effectiveness analysis

OBJECTIVE

To compare the health outcomes, costs, and incremental cost-effectiveness of universal neonatal screening for sickle cell disease (SCD) with screening targeted to African Americans.

STUDY DESIGN

A cost-effectiveness analysis was done by using a Markov simulation model that considered the costs and outcomes associated with the prevention and treatment of sepsis in those with sickle cell anemia and sickle beta(0)-thalassemia. Three strategies were compared: (1) no screening, (2) targeted screening of African Americans, and (3) universal screening for SCD.

RESULTS

In the base case analysis, targeted screening of African Americans compared with no screening cost $6709 per additional year of life saved, and universal screening compared with targeted screening cost $30,760 per additional year of life saved. In a sensitivity analysis, the cost per additional year of life saved with universal screening compared with targeted screening was positively correlated with the delivery rate of targeted screening and was inversely related to the proportion of African Americans in the population.

CONCLUSIONS

Targeted screening of African American newborns for SCD compared with no screening is always cost-effective. Universal screening compared with targeted screening always identifies more infants with disease, prevents more deaths, and is cost-effective given certain delivery rates for targeted screening and proportions of African Americans in the population.

Feline asthma syndrome in African lions (Panthera leo)

Feline asthma syndrome, previously recognized only in domestic cats, was diagnosed in three captive African lions (Panthera leo), one of which died as a result of the condition. Two of the lions displayed progressive signs for 7 yr, including severe bouts of coughing, wheezing, dyspnea, rhonchi, and tachypnea that were most severe during the spring and summer, and the third lion displayed acute signs only once. Scattered to diffuse increased interstitial markings, peribronchial cuffing, and focal atelectasis were visible in radiographs. At necropsy, multiple subpleural bullae, 2-3 cm in diameter, were scattered throughout the lung tissue. There were thick-walled bronchi and bronchioles filled with thick grayish mucus, and alveolar spaces were enlarged with severe, diffuse, banded multifocal areas of alveolar wall fibrosis. The lions had significantly elevated IgE type I immediate hypersensitivity responses to recognized aeroallergens. The captive management of lions should address the design and maintenance of allergen-free air supplies. Ventilation systems should be examined routinely and thoroughly cleaned of any residue. The frequency of examination should increase during the summer. Lions and other large cats should be routinely screened for IgE aeroallergen-specific titers, asthma cases should be treated promptly with prednisolone, and investigations of etiology should be initiated.

Efficacy of three-phase bone scans in evaluating diabetic foot ulcers

To evaluate the utility of bone scans in determining the treatment of diabetic patients with foot ulcers, a retrospective study was conducted. Medical records were reviewed for clinical signs of infection, laboratory data, and the radiologists' interpretations of imaging studies. During the study period, 34 bone scans were obtained by the treating physicians to evaluate diabetic foot ulcers. Among these, 22 of 34 bone scans were markedly confirmatory of being "consistent with osteomyelitis," 8 of 34 were moderately confirmatory ("indeterminate with regard to osteomyelitis"), and 4 of 34 were not confirmatory ("not consistent with osteomyelitis"). Of the 22 patients in the markedly confirmatory group, eight patients with clinical findings of uncontrolled infection or gangrene were treated with partial or complete amputation, whereas all others (14 patients) were treated with local wound care+/-intravenous antibiotics. Among the eight bone scans interpreted as indeterminate, three patients required partial or complete amputation, whereas the other five patients were managed with local wound care. Of the four patients with nonconfirmatory bone scans, two patients had evidence of dry gangrene and required amputation, whereas the other two patients did not have clinical evidence of infection or gangrene and were treated with local wound care. There was no significant difference in the amputation rate for patients with confirmatory, indeterminate, or nonconfirmatory bone scans for osteomyelitis (36%, 37%, and 50%, respectively) (P > 0.5). Therefore, the authors concluded that the ultimate treatment should be based on clinical indicators of the presence of uncontrolled infection or gangrene rather than on bone scan findings.

Reliability of radiographic evaluation for acromial morphology

OBJECTIVE

Bigliani's classification system of acromial morphology utilizing the standard outlet radiograph has become in accepted method for evaluating patients with rotator cuff disease. This study evaluates the interobserver and intraobserver reliability of Bigliani's classification system using observers at various levels of training.

PATIENTS AND DESIGN

Supraspinatus outlet view radiographs of 40 patients (aged 18-78 years) with shoulder pain were reviewed twice, 4 months apart, in a masked protocol by six reviewers, including two attending (fellowship-trained) shoulder surgeons, an attending musculoskeletal radiologist, an orthopedic surgery sports fellow, and two orthopedic residents (PGY-2 and PGY-5). The reviewers were given standard diagrams of the Bigliani classification system and were asked to classify each film as a type I, II, or III acromion. Interobserver reliability and intraobserver repeatability values were calculated using kappa statistic analysis (0-0.2 slight, 0.21-0.4 fair, 0.41-0.6 moderate, 0.61-0.8 substantial, and 0.8-1.0 excellent).

RESULTS AND CONCLUSION

For each of the two readings, all six observers agreed only 18% of the time. Kappa values for pairwise comparison of interobserver reliability among the six observers ranged from 0.01 to 0.75 (mean 0.35), and intraobserver repeatability ranged from 0.26 (PGY-5 resident) to 0.80 (fellowship-trained surgeon), with a mean of 0.55. Intraobserver repeatability was not significantly different for the different levels of expertise. More definitive criteria are needed to distinguish and classify the acromion.

Doxycycline compared with azithromycin for treating women with genital Chlamydia trachomatis infections: an incremental cost-effectiveness analysis

OBJECTIVE

To compare the economic consequences of doxycycline therapy with those of azithromycin therapy for women with uncomplicated cervical chlamydial infections.

DESIGN

Decision analysis in which the health outcomes, costs, and cost-effectiveness of two provider-administered treatment strategies for women with uncomplicated cervical chlamydial infections were compared: 1) initial therapy with doxycycline, 100 mg orally twice daily for 7 days (estimated cost, $5.51) and 2) initial therapy with azithromycin, 1 g orally administered as a single dose (estimated cost, $18.75).

RESULTS

Under baseline assumptions, the azithromycin strategy incurred fewer major and minor complications and was less expensive overall than the doxycycline strategy despite a higher initial cost for acquiring antibiotic agents. In univariate sensitivity analyses, the azithromycin strategy prevented more major complications but was more expensive than the doxycycline strategy when doxycycline effectiveness was greater than 0.93. In a multivariate sensitivity analysis combining 11 parameter estimates selected so that the cost-effectiveness of the doxycycline strategy would be maximized relative to that of the azithromycin strategy, the azithromycin strategy resulted in fewer complications but was more costly. The incremental cost-effectiveness was $521 per additional major complication prevented. However, if the difference in the cost of azithromycin and doxycycline decreased to $9.80, the azithromycin strategy was less expensive and more effective, even under these extreme conditions.

CONCLUSIONS

On the basis of the best available data as derived from the literature and experts, the azithromycin strategy was more cost-effective than the doxycycline strategy for women with uncomplicated cervical chlamydial infections. Despite the dominance of the azithromycin strategy over the doxycycline strategy, the adoption of the azithromycin strategy may be limited by the practical financial constraints of our currently fragmented health care system, in which the costs and benefits of preventing chlamydia sequelae are often incurred by different components of the system.

Economic analysis of roentgenogram use in the closed treatment of stable ankle fractures

Fractures of the ankle are one of the most commonly treated injuries by orthopedic surgeons. The adequacy of closed treatment of stable lateral malleolar ankle fractures is frequently assessed by repeated roentgenograms. There are no standards, nor studies, however, that provide guidelines as to the necessity of such roentgenograms. This study was designed to determine the average frequency of follow-up roentgenograms in ankle fractures treated by casting, as well as the clinical impact of these roentgenograms. The clinical radiographic data base of a university hospital was reviewed to identify all ankle fractures treated between January 1, 1992 and June 30, 1993. A total of 82 patients satisfied the study criteria of having sustained a stable ankle fracture that was treated by closed means, with sufficient clinical and radiographic follow-up to assess healing of the fracture. All patients healed their fractures at an average of 8.4 weeks (+/- 3.0 weeks), with weight-bearing initiated at 4.0 weeks (+/- 2.7 weeks). No patients developed radiographic evidence of a talar shift during treatment, and none required surgery for a failure of closed treatment. At no time did any ankle exhibit a significant change in fibular alignment relative to the initial injury films. Each patient had an average of 4.5 (+/- 2.0) radiographic studies performed throughout their treatment. This study indicates that secondary displacement of either the talus or fibula in a stable ankle fracture is very unusual. In conjunction with the generally excellent outcome for such fractures, this suggests that frequent roentgenograms are not justified on clinical grounds.(ABSTRACT TRUNCATED AT 250 WORDS)

Occupational injuries in a poor inner-city population

This study aims to characterize occupational injuries in a defined poor inner-city population in terms of demographic features, types, and circumstances of injuries, and medical and financial consequences. It is a case series drawn from a larger population-based injury registry in emergency departments that serve 17 poor census tracts in Philadelphia. Of 335 patients from the study area who had been treated at the emergency departments under study for occupational injuries, 107 could be contacted by telephone 2 to 3 years after their injuries. Interviews sought information on the patients, their employment, their injuries, and the consequences. Respondents were almost all African-American, approximately 50% male, and had a median age of 32. Approximately one third were employed in the health care industry, one fourth in the service sector (including conventional service firms, restaurants, and hotels), and the remainder in construction, retail and wholesale trade, education, transportation, and manufacturing. Major causes of injuries included overexertion, contact with sharp objects, and falls. Major types on injuries included sprain/strains and lacerations. Approximately half the respondents had missed more than 3 days of work, with 15% missing more than 1 month. Almost 40% of respondents reported persistent health problems after their injuries. Only about one quarter had received workers' compensation. We conclude that poor and minority workers are at risk of a wide range of occupational injuries, which may result in considerable lost work time and have serious medical and economic consequences. More, attention to the workplace risks of these relatively marginalized workers and more vigorous preventive interventions are needed.