Accuracy of soft-copy digital mammography versus that of screen-film mammography according to digital manufacturer: ACRIN DMIST retrospective multireader study

PURPOSE

To retrospectively compare the accuracy for cancer diagnosis of digital mammography with soft-copy interpretation with that of screen-film mammography for each digital equipment manufacturer, by using results of biopsy and follow-up as the reference standard.

MATERIALS AND METHODS

The primary HIPAA-compliant Digital Mammographic Imaging Screening Trial (DMIST) was approved by the institutional review board of each study site, and informed consent was obtained. The approvals and consent included use of data for future HIPAA-compliant retrospective research. The American College of Radiology Imaging Network DMIST collected screening mammography studies performed by using both digital and screen-film mammography in 49 528 women (mean age, 54.6 years; range, 19-92 years). Digital mammography systems from four manufacturers (Fischer, Fuji, GE, and Hologic) were used. For each digital manufacturer, a cancer-enriched reader set of women screened with both digital and screen-film mammography in DMIST was constructed. Each reader set contained all cancer-containing studies known for each digital manufacturer at the time of reader set selection, together with a subset of negative and benign studies. For each reader set, six or 12 experienced radiologists attended two randomly ordered reading sessions 6 weeks apart. Each radiologist identified suspicious findings and rated suspicion of breast cancer in identified lesions by using a seven-point scale. Results were analyzed according to digital manufacturer by using areas under the receiver operating characteristic curve (AUCs), sensitivity, and specificity for soft-copy digital and screen-film mammography. Results for Hologic digital are not presented owing to the fact that few cancer cases were available. The implemented design provided 80% power to detect average AUC differences of 0.09, 0.08, and 0.06 for Fischer, Fuji, and GE, respectively.

RESULTS

No significant difference in AUC, sensitivity, or specificity was found between Fischer, Fuji, and GE soft-copy digital and screen-film mammography. Large reader variations occurred with each modality.

CONCLUSION

No statistically significant differences were found between soft-copy digital and screen-film mammography for Fischer, Fuji, and GE digital mammography equipment.

Early invasive cervical cancer: CT and MR imaging in preoperative evaluation - ACRIN/GOG comparative study of diagnostic performance and interobserver variability

PURPOSE

To retrospectively compare diagnostic performance and interobserver variability for computed tomography (CT) and magnetic resonance (MR) imaging in the pretreatment evaluation of early invasive cervical cancer, with surgical pathologic findings as the reference standard.

MATERIALS AND METHODS

This HIPAA-compliant study had institutional review board approval and informed consent for evaluation of preoperative CT (n = 146) and/or MR imaging (n = 152) studies in 156 women (median age, 43 years; range, 22-81 years) from a previous prospective multicenter American College of Radiology Imaging Network and Gynecologic Oncology Group study of 172 women with biopsy-proved cervical cancer (clinical stage > or = IB). Four radiologists (experience, 7-15 years) interpreted the CT scans, and four radiologists (experience, 12-20 years) interpreted the MR studies retrospectively. Tumor visualization and detection of parametrial invasion were assessed with receiver operating characteristic curves (with P < or = .05 considered to indicate a significant difference). Descriptive statistics for staging and kappa statistics for reader agreement were calculated. Surgical pathologic findings were the reference standard.

RESULTS

For CT and MR imaging, respectively, multirater kappa values were 0.26 and 0.44 for staging, 0.16 and 0.32 for tumor visualization, and -0.04 and 0.11 for detection of parametrial invasion; for advanced stage cancer (> or =IIB), sensitivities were 0.14-0.38 and 0.40-0.57, positive predictive values (PPVs) were 0.38-1.00 and 0.32-0.39, specificities were 0.84-1.00 and 0.77-0.80, and negative predictive values (NPVs) were 0.81-0.84 and 0.83-0.87. MR imaging was significantly better than CT for tumor visualization (P < .001) and detection of parametrial invasion (P = .047).

CONCLUSION

Reader agreement was higher for MR imaging than for CT but was low for both. MR imaging was significantly better than CT for tumor visualization and detection of parametrial invasion. The modalities were similar for staging, sharing low sensitivity and PPV but relatively high NPV and specificity.

Cancer yield of mammography, MR, and US in high-risk women: prospective multi-institution breast cancer screening study

PURPOSE

To prospectively determine cancer yield, callback and biopsy rates, and positive predictive value (PPV) of mammography, magnetic resonance (MR) imaging, and ultrasonography (US) in women at high risk for breast cancer.

MATERIALS AND METHODS

The study was approved by the institutional review board and was HIPAA compliant, and informed consent was obtained. We conducted a prospective pilot study of screening mammography, MR, and US in asymptomatic women 25 years of age or older who were genetically at high risk, defined as BRCA1/BRCA2 carriers or with at least a 20% probability of carrying a BRCA1/BRCA2 mutation. All imaging modalities were performed within 90 days of each other. Data were analyzed by using exact confidence intervals (CIs) and the McNemar test.

RESULTS

A total of 195 women were enrolled in this study over a 6-month period, and 171 completed all study examinations (mammography, US, and MR). Average age of the 171 participants was 46 years +/- 10.2 (standard deviation). Sixteen biopsies were performed and six cancers were detected, for an overall 3.5% cancer yield. MR enabled detection of all six cancers; mammography, two; and US, one. The diagnostic yields for each test were 3.5% for MR, 0.6% for US, and 1.2% for mammography. MR, US, and mammography findings prompted biopsy in 8.2%, 2.3%, and 2.3% of patients, respectively. None of the pairwise comparisons were statistically significant. The PPV of biopsies performed as a result of MR was 43%.

CONCLUSION

Screening MR imaging had a higher biopsy rate but helped detect more cancers than either mammography or US. US had the highest false-negative rate compared with mammography and MR, enabling detection of only one in six cancers in high-risk women.

Factors influencing performance of MRI in newly diagnosed breast cancer patients

604

Background: There is wide variation in the use of breast MRI to screen the contralateral breast in patients recently diagnosed with unilateral breast cancer. Some centers routinely scan all patients with a recent diagnosis of breast cancer while others restrict breast MRI to patients with dense breast tissue, patients who are pre-menopausal, or patients with infiltrating lobular carcinoma. This international study evaluated performance measures of MRI by mammographic density, menopausal status and type of index cancer. Methods: 969 women from 21 sites with a recent diagnosis of unilateral breast cancer and a negative mammogram and CBE of the contralateral breast underwent breast MRI. Presence of breast cancer in the contralateral breast was determined by cancer positive breast biopsy within 12 months after study entry. Performance measures of breast MRI (cancer yield, sensitivity, specificity, negative predictive value, positive predictive value, biopsy rate) were compared between participant subsets defined by mammographic density (fatty vs dense), menopausal status (pre/peri menopausal vs post menopausal) and type of index cancer (invasive vs in situ and lobular vs non-lobular). Results: Performance measures of breast MRI were not influenced by breast density or index cancer histology. Cancer yield of MRI in dense breast women was 3% in both fatty and dense breasted women. Although cancer yield and sensitivity of MRI did not vary based on menopausal status, specificity was significantly higher among post-menopausal women in comparison to pre- or peri-menopausal women (p-value=0.002) as was positive biopsy rate (p-value 0.009). Conclusions: Performance of MRI in screening the contralateral breast in the newly diagnosed breast cancer patient is not influenced by breast density or index cancer histology. Specificity and positive biopsy rate are higher in post-menopausal women, which may be related to hormonal influences on breast tissue enhancement.

[Table: see text]

MRI evaluation of the contralateral breast in women with recently diagnosed breast cancer

BACKGROUND

Even after careful clinical and mammographic evaluation, cancer is found in the contralateral breast in up to 10% of women who have received treatment for unilateral breast cancer. We conducted a study to determine whether magnetic resonance imaging (MRI) could improve on clinical breast examination and mammography in detecting contralateral breast cancer soon after the initial diagnosis of unilateral breast cancer.

METHODS

A total of 969 women with a recent diagnosis of unilateral breast cancer and no abnormalities on mammographic and clinical examination of the contralateral breast underwent breast MRI. The diagnosis of MRI-detected cancer was confirmed by means of biopsy within 12 months after study entry. The absence of breast cancer was determined by means of biopsy, the absence of positive findings on repeat imaging and clinical examination, or both at 1 year of follow-up.

RESULTS

MRI detected clinically and mammographically occult breast cancer in the contralateral breast in 30 of 969 women who were enrolled in the study (3.1%). The sensitivity of MRI in the contralateral breast was 91%, and the specificity was 88%. The negative predictive value of MRI was 99%. A biopsy was performed on the basis of a positive MRI finding in 121 of the 969 women (12.5%), 30 of whom had specimens that were positive for cancer (24.8%); 18 of the 30 specimens were positive for invasive cancer. The mean diameter of the invasive tumors detected was 10.9 mm. The additional number of cancers detected was not influenced by breast density, menopausal status, or the histologic features of the primary tumor.

CONCLUSIONS

MRI can detect cancer in the contralateral breast that is missed by mammography and clinical examination at the time of the initial breast-cancer diagnosis. (ClinicalTrials.gov number, NCT00058058 [ClinicalTrials.gov].).

Meta-analysis of diagnostic and screening test accuracy evaluations: methodologic primer

OBJECTIVE

Interest in evidence-based diagnosis is growing rapidly as diagnostic and screening techniques proliferate. In this article we provide an overview of systematic reviews of diagnostic performance and discuss in detail statistical methods for the most common variant of the problem: meta-analysis of studies in which a pair of estimates of sensitivity and specificity is reported. The need to account for possible variations in threshold for test positivity across studies led to the formulation of the Summary ROC (SROC) curve method. We discuss graphical and model-based ways to estimate, summarize, and compare SROC curves, and we present an example from a meta-analysis of data on techniques for staging cervical cancer. We also present a brief survey of the methodologic literature for addressing heterogeneity, correlated data, multiple thresholds per study, and systematic reviews of ROC studies. We conclude with a discussion of the significant methodologic challenges that continue to face investigators in this area of diagnostic medicine research.

CONCLUSION

Systematic reviews of diagnostic performance are a rigorous approach to examining and synthesizing evidence in the evaluation of diagnostic and screening tests. The information from such reviews is needed by clinicians, health policy makers, researchers in diagnostic medicine, developers of diagnostic techniques, and the general public. However, despite progress in study quality and reporting and in methodologic development, major challenges confront investigators undertaking these reviews.

Role of Imaging in Pretreatment Evaluation of Early Invasive Cervical Cancer: Results of the Intergroup Study American College of Radiology Imaging Network 6651–Gynecologic Oncology Group 183

Purpose

To compare magnetic resonance imaging (MRI) and computed tomography (CT) with each other and to International Federation of Gynecology and Obstetrics (FIGO) clinical staging in the pretreatment evaluation of early invasive cervical cancer, using surgicopathologic findings as the reference standard.

Patients and Methods

This prospective multicenter clinical study was conducted by the American College of Radiology Imaging Network and the Gynecologic Oncology Group from March 2000 to November 2002; 25 United States health centers enrolled 208 consecutive patients with biopsy-confirmed cervical cancer of FIGO stage ≥ IB who were scheduled for surgery based on clinical assessment. Patients underwent FIGO clinical staging, helical CT, and MRI. Surgicopathologic findings constituted the reference standard for statistical analysis.

Results

Complete data were available for 172 patients; surgicopathologic findings were consistent with FIGO stages IA to IIA in 76% and stage ≥ IIB in 21%. For the detection of advanced stage (≥ IIB), sensitivity was poor for FIGO clinical staging (29%), CT (42%), and MRI (53%); specificity was 99% for FIGO clinical staging, 82% for CT, and 74% for MRI; and negative predictive value was 84% for FIGO clinical staging, 84% for CT, and 85% for MRI. MRI (area under the receiver operating characteristic curve [AUC], 0.88) was significantly better than CT (AUC, 0.73) for detecting cervical tumors (P = .014). For 85% of patients, FIGO clinical staging forms were submitted after MRI and/or CT was performed.

Conclusion

CT and MRI performed similarly; both had lower staging accuracy than in prior single-institution studies. Accuracy of FIGO clinical staging was higher than previously reported. The temporal data suggest that FIGO clinical staging was influenced by CT and MRI findings.

Utilization of diagnostic studies in the pretreatment evaluation of invasive cervical cancer in the United States: results of intergroup protocol ACRIN 6651/GOG 183

PURPOSE

To review the current utilization of diagnostic tests prescribed by the International Federation of Gynecology and Obstetrics (FIGO) clinical staging guidelines in the pretreatment work-up of invasive cervical cancer, and to compare the data with those of previous patterns of care studies.

PATIENTS AND METHODS

This interdisciplinary American College of Radiology Imaging Network/Gynecologic Oncology Group prospective clinical trial was conducted between March 1, 2000, and November 11, 2002. Twenty-five participating institutions, all from the United States, enrolled a total of 208 patients. Only patients scheduled for surgery with biopsy-confirmed cervical cancer of clinical FIGO stage IB or higher were eligible. The patterns of care data analysis was based on 197 patients who met all inclusion criteria. The conventional FIGO-recommended tests used for pre-enrollment FIGO clinical stage classification were at the discretion of the treating physician; overall frequency of use was tabulated for each test.

RESULTS

Use of cystoscopy (8.1%) and sigmoidoscopy or proctoscopy (8.6%) was significantly lower than in 1988 to 1989 (P < .0001 in each instance). Intravenous urography was used in only 1% of patients as compared with 42% in 1988 to 1989 and 91% in 1983. No patient included in the data analysis had barium enema or lymphangiography. Only 26.9% of patients had examination under anesthesia for FIGO clinical staging.

CONCLUSION

There is a large discrepancy between the diagnostic tests recommended by FIGO and the actual tests used for cervical cancer staging, suggesting a need to reassess the relevance of the FIGO guidelines to current clinical practice in the United States.

Diagnostic performance of digital versus film mammography for breast-cancer screening

BACKGROUND

Film mammography has limited sensitivity for the detection of breast cancer in women with radiographically dense breasts. We assessed whether the use of digital mammography would avoid some of these limitations.

METHODS

A total of 49,528 asymptomatic women presenting for screening mammography at 33 sites in the United States and Canada underwent both digital and film mammography. All relevant information was available for 42,760 of these women (86.3 percent). Mammograms were interpreted independently by two radiologists. Breast-cancer status was ascertained on the basis of a breast biopsy done within 15 months after study entry or a follow-up mammogram obtained at least 10 months after study entry. Receiver-operating-characteristic (ROC) analysis was used to evaluate the results.

RESULTS

In the entire population, the diagnostic accuracy of digital and film mammography was similar (difference between methods in the area under the ROC curve, 0.03; 95 percent confidence interval, -0.02 to 0.08; P=0.18). However, the accuracy of digital mammography was significantly higher than that of film mammography among women under the age of 50 years (difference in the area under the curve, 0.15; 95 percent confidence interval, 0.05 to 0.25; P=0.002), women with heterogeneously dense or extremely dense breasts on mammography (difference, 0.11; 95 percent confidence interval, 0.04 to 0.18; P=0.003), and premenopausal or perimenopausal women (difference, 0.15; 95 percent confidence interval, 0.05 to 0.24; P=0.002).

CONCLUSIONS

The overall diagnostic accuracy of digital and film mammography as a means of screening for breast cancer is similar, but digital mammography is more accurate in women under the age of 50 years, women with radiographically dense breasts, and premenopausal or perimenopausal women. (ClinicalTrials.gov number, NCT00008346.)

Added cancer yield of MRI in screening the contralateral breast of women recently diagnosed with breast cancer: results from the International Breast Magnetic Resonance Consortium (IBMC) trial

OBJECTIVE

To estimate the added cancer yield of magnetic resonance imaging (MRI) over mammography in the contralateral breast of patients with a recent diagnosis of breast cancer.

METHODS

We conducted a prospective, international study of mammography and MRI in women with a recent diagnosis of unilateral breast cancer. Each subject received a mammogram, clinical breast exam (CBE), and MRI of the unaffected breast within a 90 day time period. Definitive diagnosis of suspicious findings was determined through biopsy and central pathology review.

RESULTS

Of the 103 eligible women included in study analyses, MRI detected 4 cancers in the contralateral breast while mammography detected none. MRI resulted in 12% (95% CI, 6%-20%) of women recommended for biopsy and 10% of women undergoing additional biopsy. The added cancer yield of MRI was 4% (95% CI, 1%-10%) and the positive predictive value of an abnormal MRI was 33% (95% CI, 10%-65%). Forty percent (4/10) of the biopsies performed based on the MRI recommendation were positive for malignancy.

CONCLUSION

In women with a recent breast cancer diagnosis, approximately 4% will have an otherwise occult invasive breast cancer detected in the opposite breast by MRI alone.

Screening women at high risk for breast cancer with mammography and magnetic resonance imaging

BACKGROUND

The authors compared the performance of screening mammography versus magnetic resonance imaging (MRI) in women at genetically high risk for breast cancer.

METHODS

The authors conducted an international prospective study of screening mammography and MRI in asymptomatic, genetically high-risk women age >/= 25 years. Women with a history of breast cancer were eligible for a contralateral screening if they had been diagnosed within 5 years or a bilateral screening if they had been diagnosed > 5 years previously. All examinations (MRI, mammography, and clinical breast examination [CBE]) were performed within 90 days of each other.

RESULTS

In total, 390 eligible women were enrolled by 13 sites, and 367 women completed all study examinations. Imaging evaluations recommended 38 biopsies, and 27 biopsies were performed, resulting in 4 cancers diagnosed for an overall 1.1% cancer yield (95% confidence interval [95%CI], 0.3-2.8%). MRI detected all four cancers, whereas mammography detected one cancer. The diagnostic yield of mammography was 0.3% (95%CI, 0.01-1.5%). The yield of cancer by MRI alone was 0.8% (95%CI, - 0.3-2.0%). The biopsy recommendation rates for MRI and mammography were 8.5% (95%CI, 5.8-11.8%) and 2.2% (95%CI, 0.1-4.3%).

CONCLUSIONS

Screening MRI in high-risk women was capable of detecting mammographically and clinically occult breast cancer. Screening MRI resulted in 22 of 367 of women (6%) who had negative mammogram and negative CBE examinations undergoing biopsy, resulting in 3 additional cancers detected. MRI also resulted in 19 (5%) false-positive outcomes, which resulted in benign biopsies.

Error reporting and disclosure systems: views from hospital leaders

CONTEXT

The Institute of Medicine has recommended establishing mandatory error reporting systems for hospitals and other health settings.

OBJECTIVE

To examine the opinions and experiences of hospital leaders with state reporting systems.

DESIGN AND SETTING

Survey of chief executive and chief operating officers (CEOs/COOs) from randomly selected hospitals in 2 states with mandatory reporting and public disclosure, 2 states with mandatory reporting without public disclosure, and 2 states without mandatory systems in 2002-2003.

MAIN OUTCOME MEASURES

Perceptions of the effects of mandatory systems on error reporting, likelihood of lawsuits, and overall patient safety; attitudes regarding release of incident reports to the public; and likelihood of reporting incidents to the state or to the affected patient based on hypothetical clinical vignettes that varied the type and severity of patient injury.

RESULTS

Responses were received from 203 of 320 hospitals (response rate = 63%). Most CEOs/COOs thought that a mandatory, nonconfidential system would discourage reporting of patient safety incidents to their hospital's own internal reporting system (69%) and encourage lawsuits (79%) while having no effect or a negative effect on patient safety (73%). More than 80% felt that the names of both the hospital and the involved professionals should be kept confidential, although respondents from states with mandatory public disclosure systems were more willing than respondents from the other states to release the hospital name (22% vs 4%-6%, P = .005). Based on the vignettes, more than 90% of hospital leaders said their hospital would report incidents involving serious injury to the state, but far fewer would report moderate or minor injuries, even when the incident was of sufficient consequence that they would tell the affected patient or family.

CONCLUSIONS

Most hospital leaders expressed substantial concerns about the impact of mandatory, nonconfidential reporting systems on hospital internal reporting, lawsuits, and overall patient safety. While hospital leaders generally favor disclosure of patient safety incidents to involved patients, fewer would disclose incidents involving moderate or minor injury to state reporting systems.

Meta-analysis of diagnostic test accuracy assessment studies with varying number of thresholds

Current meta-analytic methods for diagnostic test accuracy are generally applicable to a selection of studies reporting only estimates of sensitivity and specificity, or at most, to studies whose results are reported using an equal number of ordered categories. In this article, we propose a new meta-analytic method to evaluate test accuracy and arrive at a summary receiver operating characteristic (ROC) curve for a collection of studies evaluating diagnostic tests, even when test results are reported in an unequal number of nonnested ordered categories. We discuss both non-Bayesian and Bayesian formulations of the approach. In the Bayesian setting, we propose several ways to construct summary ROC curves and their credible bands. We illustrate our approach with data from a recently published meta-analysis evaluating a single serum progesterone test for diagnosing pregnancy failure.

Race and Gender Disparities in Rates of Cardiac Revascularization

BACKGROUND

Numerous studies have documented substantial differences by race and gender in the use of coronary artery bypass graft surgery and percutaneous coronary angioplasty. However, few studies have examined whether these differences reflect problems in quality of care.

METHOD

We selected a random sample stratified by gender, race, and income of 5026 Medicare beneficiaries aged 65 to 75 who underwent inpatient coronary angiography during 1991 to 1992 in 1 of 5 states. We compared the frequency of 2 problems in quality by race and gender: underuse or the failure to receive a clinically indicated revascularization procedure and receipt of revascularization when it was not clinically indicated. We used 2 independent sets of criteria developed by the RAND Corporation and the American College of Cardiology/American Hospital Association (ACC/AHA). We also examined survival of the cohort through March 31, 1994.

RESULTS

Revascularization procedures were clinically indicated more frequently among whites than blacks and among men than women. Failure to receive revascularization when it was indicated was more common among blacks than among whites (40% vs. 23-24%, depending on the criteria, both P<0.001) but similar among men and women (25% vs. 22-24%, P>0.05). Racial disparities remained similar after adjusting for patient and hospital characteristics. Among patients rated inappropriate, use of procedures was greater for whites than blacks using RAND criteria (10.5% vs. 5.8%, P<0.01) and greater for men than for women (14.2% vs. 5.3% by RAND criteria, P=0.001; 8.2% vs. 4.0%% by ACC/AHA criteria, P=0.04). After multivariate adjustment, the disparities for race and gender remained similar and were statistically significant using RAND criteria. Mortality rates tended to validate our appropriateness criteria for underuse.

CONCLUSIONS

Racial differences in procedure use reflect higher rates of clinical appropriateness among whites, greater underuse among blacks, and more frequent revascularization when it was not clinically indicated among whites. Underuse is associated with higher mortality. In contrast, men had higher rates of clinical appropriateness and were more likely to receive revascularization when it was not clinically indicated. There was no evidence of greater underuse among women.

Primer on Multiple Regression Models for Diagnostic Imaging Research

This article provides an introduction to multiple regression analysis and its application in diagnostic imaging research. We begin by examining why multiple regression models are needed in the evaluation of diagnostic imaging technologies. We then examine the broad categories of available models, notably multiple linear regression models for continuous outcomes and logistic regression models for binary outcomes. The purpose of this article is to elucidate the scientific logic, meaning, and interpretation of multiple regression models by using examples from the diagnostic imaging literature.

Adherence to practice guidelines: the role of specialty society guidelines

BACKGROUND

Physician adherence to guidelines is often poor, but the reasons have not been completely studied. We investigated whether physician adherence to guidelines for percutaneous transluminal coronary angioplasty (PTCA) and coronary artery bypass grafting (CABG) varied by source, development methods, or the extent of their evidence-base.

METHODS AND RESULTS

We assessed adherence to guidelines developed by the American College of Cardiology/American Heart Association (ACC/AHA) for PTCA (1988 and 1993) and for CABG (1990) and guidelines developed by RAND for PTCA and CABG in 1990. We randomly sampled patients on Medicare who were undergoing coronary angiography in 5 states in 1991 and 1992, extracting clinical and laboratory data from medical records and using computer programs to classify the appropriateness of each procedure. A total of 543 PTCA and 676 CABG procedures were studied. By use of the 1988 ACC/AHA guidelines, 30% of PTCAs were rated class III (inappropriate), whereas 24% were class III by use of the 1993 guidelines. Only 1.5% of CABG procedures were class III with ACC/AHA guidelines. By use of RAND guidelines, 12% of PTCA and 9% of CABG procedures were classified as inappropriate.

CONCLUSIONS

Adherence to guidelines is higher when the recommendations are supported by evidence from randomized clinical trials (CABG). The credibility of the source and familiarity with the guidelines do not ensure compliance. When evidence is lacking, as with PTCA at the time of this study, guideline recommendations may lag behind appropriate changes in clinical practice. More frequent revisions coupled with on-line access have the potential to make guidelines more useful.

Racial differences in cardiac revascularization rates: does "overuse" explain higher rates among white patients?

BACKGROUND

Coronary artery bypass graft (CABG) surgery and percutaneous transluminal coronary angioplasty (PTCA) are well-established treatments for symptomatic coronary artery disease. Previous studies have documented racial differences in rates of use of these cardiac revascularization procedures. Other studies suggest that these procedures are overused: that is, they are done for patients with clinically inappropriate indications.

OBJECTIVE

To test the hypothesis that the higher rate of cardiac revascularization among white patients is associated with a higher prevalence of overuse (revascularization for clinically inappropriate indications) among white patients than among African-American patients.

DESIGN

Observational cohort study using Medicare claims and medical record review.

SETTING

173 hospitals in five U.S. states.

PARTICIPANTS

A stratified, weighted, random sample of 3960 Medicare beneficiaries who underwent coronary angiography during 1991 and 1992; 1692 of these patients underwent 1711 revascularization procedures within 90 days.

MEASUREMENTS

The proportion of CABG and PTCA procedures rated appropriate, uncertain, and inappropriate according to RAND criteria, and the multivariate odds of undergoing inappropriate revascularization among African-American patients and white patients.

RESULTS

After angiography, rates of PTCA (23% vs. 19%) and CABG surgery (29% vs. 17%) were significantly higher among white patients than among African-American patients. The respective rates of inappropriate PTCA and CABG surgery were 14% and 10%. Among the study states, rates of inappropriate use ranged from 4% to 24% for PTCA and 0% to 14% for CABG surgery. White patients were more likely than African-American patients to receive inappropriate PTCA (15% vs. 9%; difference, 6 percentage points [95% CI, -0.4 to 12.7 percentage points]), and difference by race was statistically significant among men (20% vs. 8%; difference, 12 percentage points [CI, 1.2 to 21.7 percentage points]). Rates of inappropriate CABG surgery did not differ by race (10% in both groups).

CONCLUSIONS

Among a large and diverse sample of Medicare beneficiaries in five U.S. states, overuse of PTCA was greater among white men than among other groups, but this difference did not fully account for racial disparities in revascularization. Overuse of cardiac revascularization varied significantly by geographic region.

Effects of angiotensin-converting enzyme inhibitors and digoxin on health outcomes of very old patients with heart failure. SAGE Study Group. Systematic Assessment of Geriatric drug use via Epidemiology

BACKGROUND

Randomized trials have shown that angiotensin-converting enzyme (ACE) inhibitors reduce mortality and morbidity, and improve symptoms and exercise tolerance in selected patients with congestive heart failure (CHF). There is, however, no evidence on the effectiveness of ACE inhibitors in the typical, very old and frail patients with CHF.

OBJECTIVE

To compare the effects of ACE inhibitors and digoxin on 1-year mortality, morbidity, and physical function among patients aged 85 years.

METHODS

We conducted a retrospective cohort study using the SAGE database, a long-term care database linking patient information with drug utilization data. Among 64637 patients with CHF admitted to all nursing homes in 5 states between 1992 and 1995, we identified 19492 patients taking either an ACE inhibitor (n = 4911) or digoxin (n = 14890). Record of date of death was derived from Medicare enrollment files, and we used the part A Medicare files to identify hospital admissions and discharge diagnoses. As a measure of physical function, we used a scale for activities of daily living performance. The effect of ACE inhibitors was estimated using Cox proportional hazards models with digoxin users as the reference group.

RESULTS

The overall mortality rate among ACE inhibitor recipients was more than 10% less than that of digoxin users (relative rate, 0.89; 95% confidence interval, 0.83-0.95). Mortality was equally reduced regardless of concomitant cardiovascular conditions and baseline physical function. Treatment with ACE inhibitors was associated with a tendency toward reduced hospital admissions that was more evident among patients with greater functional impairment. The adjusted relative rate for hospitalization for any reason was 0.96 (95% confidence interval, 0.91-1.01). The rate of functional decline was greatly reduced among ACE inhibitor recipients (relative rate, 0.74; 95% confidence interval, 0.69-0.80), and this effect was consistent and independent of background comorbidity and baseline physical function.

CONCLUSIONS

These data suggest that survival and functional benefits of ACE inhibitor therapy extend to patients with CHF 85 years and older, and mostly women, both systematically underrepresented in randomized trials. Alternatively, digoxin has a detrimental effect in this population.

Generalized estimating equations for ordinal categorical data: arbitrary patterns of missing responses and missingness in a key covariate

We propose methods for regression analysis of repeatedly measured ordinal categorical data when there is nonmonotone missingness in these responses and when a key covariate is missing depending on observables. The methods use ordinal regression models in conjunction with generalized estimating equations (GEEs). We extend the GEE methodology to accommodate arbitrary patterns of missingness in the responses when this missingness is independent of the unobserved responses. We further extend the methodology to provide correction for possible bias when missingness in knowledge of a key covariate may depend on observables. The approach is illustrated with the analysis of data from a study in diagnostic oncology in which multiple correlated receiver operating characteristic curves are estimated and corrected for possible verification bias when the true disease status is missing depending on observables.

Ambulatory blood pressure responses and the circumplex model of mood: a 4-day study

BACKGROUND

The relation between mood or emotions and concurrent ambulatory blood pressure responses holds both fundamental and clinical interest.

METHODS

The primary sample consisted of 69 normotensive or borderline hypertensive but otherwise healthy adult males. The validation sample consisted of 85 healthy male undergraduate college students. Both samples underwent half-hourly 24-hour ambulatory blood pressure measurements on four separate workdays, 1 week apart. At each ambulatory measurement, subjects recorded their behavior, environment, and mood. The circular mood scale, a circular visual analogue scale based on the circumplex model of mood, was used to reflect the totality of a participant's affective state space. Longitudinal random effects regression models were applied in the data analysis.

RESULTS

The results for both samples were quite similar. Sleep and posture had the greatest influence on ambulatory blood pressure and heart rate. The effects of the environmental setting, social setting, and consumption were modest but statistically significant. Independent of these covariates, mood exerted a significant effect on blood pressure and heart rate. Relative to the "mellow" default category, blood pressure increased both for "anxious/annoyed" and "elated/happy" and decreased during "disengaged/sleepy" mood. The range of mood-related blood pressure estimates was 6.0/3.7 mm Hg.

CONCLUSIONS

The pattern of blood pressure responses suggests that they were related to the degree of engagement of a mood rather than the degree of unpleasantness. The hypothesis that posits that negative affect-related cardiovascular reactivity mediates the observed correlation between negative affect and disease risk should be reconsidered.

Characteristics of the SAGE database: a new resource for research on outcomes in long-term care. SAGE (Systematic Assessment of Geriatric drug use via Epidemiology) Study Group

BACKGROUND

Because there is a lack of databases specific to long-term care, standardized assessments of nursing home residents are seen as a potential new resource for studying an important but neglected population. We describe the design and principal population characteristics of the first integrated database combining detailed clinical information and administrative claims data.

METHODS

We studied nearly 300,000 residents admitted between 1992 and 1994 to all Medicare/Medicaid certified nursing homes of five U.S. states (Kansas, Maine, Mississippi, New York, and South Dakota). The database crosslinks: (a) Resident Data: over 350 items (demographic, diagnostic, clinical, and treatments) collected with the Minimum Data Set; (b) Drug Data: brand name, dosage route, and frequency of administration for all drugs consumed by each resident; (c) Medicare Data: eligibility and inpatient hospital claims; (d) Facilities Data: structural and staffing information on nursing homes; and (e) Country Data: information on population, health professions and facility data, and economic parameters.

RESULTS

Ninety-two percent of the residents were aged 65 years and older. Residents were predominantly white (85%) and female (72%). The average number of medical diagnoses was above three, and residents were receiving an average of six medications. Sixty-five percent of residents had at least one hospital claim following the initial assessment, most commonly related to cardiovascular diseases and metabolic disorders. Fifty-five percent of the facilities were for-profit and 33% were of small size. Quality indicators and staffing level varied significantly by state.

CONCLUSIONS

The SAGE (Systematic Assessment of Geriatric drug use via Epidemiology) database provides a unique resource to study the relation between treatments received and outcomes experienced, particularly functional and health services outcomes, that have not been possible before in very old, frail people.

The stability of preferences for life-sustaining care among persons with AIDS in the Boston Health Study

BACKGROUND

Clinicians recognize the importance of eliciting patient preferences for life-sustaining care, yet little is known about the stability of those preferences for patients with serious disease.

OBJECTIVES

To examine the stability of preferences for life-sustaining care among persons with AIDS and to assess factors associated with changes in preferences.

DESIGN

Two patient surveys and medical record reviews, administered four months apart in 1990-1991.

SETTING

Three health care settings in Boston.

PATIENTS

252 of 505 eligible persons with AIDS who participated in both baseline and follow-up surveys.

MAIN OUTCOME MEASURES

A single question assessing desire for cardiac resuscitation and a scale of preferences for life-extending treatment conditional on hypothetical health states.

RESULTS

Approximately one-fourth of the respondents changed their minds about life-sustaining care during a four-month period. Of patients who initially desired cardiac resuscitation, 23% decided to forego it four months later, and of those who initially said they would decline care, 34% later said they would accept it. Of those who initially desired any of the life-extending treatments, 25% decided to forego them four months later, and of those who initially said they would decline life-extending care, 24% later said they would accept some treatment. Patients reporting changes in physical function, pain, or suicide ideation were more likely to modify their desires to be resuscitated (all p< or =0.05). Patients lacking an advance directive, not completing high school, or becoming more severely ill were more likely to change their preferences on the Life Extension scale (p< or =0.05). Patients who discussed their preferences with at least one physician were just as likely as others to change desires for cardiac resuscitation. Age, gender, race, emotional health, clinical severity, social support, and site of care were not significant correlates of change for either measure.

CONCLUSIONS

Health care providers should periodically reassess preferences for life-sustaining care, particularly for patients with progressive disease, given the instability in patient preferences. However, predictors of instability may vary with how preferences are measured. In particular, changes in health status may be related to instability of preferences for certain types of treatments.

Management of pain in elderly patients with cancer. SAGE Study Group. Systematic Assessment of Geriatric Drug Use via Epidemiology

CONTEXT

Cancer pain can be relieved with pharmacological agents as indicated by the World Health Organization (WHO). All too frequently pain management is reported to be poor.

OBJECTIVE

To evaluate the adequacy of pain management in elderly and minority cancer patients admitted to nursing homes.

DESIGN

Retrospective, cross-sectional study.

SETTING

A total of 1492 Medicare-certified and/or Medicaid-certified nursing homes in 5 states participating in the Health Care Financing Administration's demonstration project, which evaluated the implementation of the Resident Assessment Instrument and its Minimum Data Set.

STUDY POPULATION

A group of 13 625 cancer patients aged 65 years and older discharged from the hospital to any of the facilities from 1992 to 1995. Data were from the multilinked Systematic Assessment of Geriatric Drug Use via Epidemiology (SAGE) database.

MAIN OUTCOME MEASURES

Prevalence and predictors of daily pain and of analgesic treatment. Pain assessment was based on patients' report and was completed by a multidisciplinary team of nursing home personnel that observed, over a 7-day period, whether each resident complained or showed evidence of pain daily.

RESULTS

A total of 4003 patients (24%, 29%, and 38% of those aged > or =85 years, 75 to 84 years, and 65 to 74 years, respectively) reported daily pain. Age, gender, race, marital status, physical function, depression, and cognitive status were all independently associated with the presence of pain. Of patients with daily pain, 16% received a WHO level 1 drug, 32% a WHO level 2 drug, and only 26% received morphine. Patients aged 85 years and older were less likely to receive morphine or other strong opiates [corrected] than those aged 65 to 74 years (13% vs 38%, respectively). More than a quarter of patients (26%) in daily pain did not receive any analgesic agent. Patients older than 85 years in daily pain were also more likely to receive no analgesia (odds ratio [OR], 1.40; 95% confidence interval [CI], 1.13-1.73). Other independent predictors of failing to receive any analgesic agent were minority race (OR, 1.63; 95% CI, 1.18-2.26 for African Americans), low cognitive performance (OR, 1.23; 95% CI, 1.05-1.44), and the number of other medications received (OR, 0.65; 95% CI, 0.5-0.84 for 11 or more medications).

CONCLUSIONS

Daily pain is prevalent among nursing home residents with cancer and is often untreated, particularly among older and minority patients.

Hierarchical polytomous regression models with applications to health services research

The analysis of variations is an important area of interest in health services and outcomes research and has two main goals: to identify and quantify variability across units, such as geographic regions or health care providers, in terms of procedure utilization and outcomes, and to explore the links between process, such as regional or hospital practice patterns, and outcomes, such as patient mortality and functional status. Hierarchical regression models are well suited for this type of analysis. In this paper we formulate a hierarchical polytomous regression model and apply it to the analysis of variations in the utilization of alternative cardiac procedures in a national cohort of elderly Medicare patients who had an acute myocardial infarction during 1987. The model is designed to accommodate clustered multinomial data with covariate vectors available on individual cases and on clusters. We present a Bayesian approach to fitting and checking the model using simulated values from the posterior distribution of the parameters. The simulation algorithms are based on Gibbs sampling in combination with Metropolis steps. Using the hierarchical polytomous regression model, we examine how the rates of cardiac procedures depend on patient-level characteristics, including age, gender and race, and whether there exist interstate differences and regional patterns in the use of these procedures.

First-episode major depressive and dysthymic disorder in childhood: clinical and sociodemographic factors in recovery

OBJECTIVE

To characterize the temporal pattern of depressive disorder in childhood, the first episode of depression was examined, focusing on recovery and its baseline predictors.

METHOD

The sample includes 112 clinically referred 8- to 13-year-olds with first-episode major depressive or dysthymic disorder participating in a naturalistic follow-up study. Psychiatric diagnoses were based on standardized interviews and operational criteria. Recovery was modeled by multivariate procedures using baseline clinical and demographic predictors.

RESULTS

Recovery rates were 86% and 7% for major depression and dysthymia, respectively, 2 years after onset. Median duration of major depression was 9 months and was predicted only by underlying dysthymia. Median duration of dysthymic disorder was 3.9 years and was predicted only by comorbid externalizing disorder. In post hoc analyses, no positive treatment effects were detected.

CONCLUSIONS

First-episode depression in youths is persistent, it generally appears to run its own course, and its naturalistic treatment requires scrutiny. However, because comorbid externalizing disorder apparently affects duration of dysthymia, intervention for behavior problems may shorten this type of depression.

The Boston AIDS Survival Score (BASS): a multidimensional AIDS severity instrument

OBJECTIVES

This study developed a new acquired immunodeficiency syndrome (AIDS) severity system by including diagnostic, physiological, functional, and sociodemographic factors predictive of survival.

METHODS

Three-hundred five persons with AIDS in Boston were interviewed; their medical records were reviewed and vital status ascertained.

RESULTS

Overall median (+/- SD) survival for the cohort from the first interview until death was 560 +/- 14.4 days. The best model for predicting survival, the Boston AIDS Survival Score, included the Justice score (stage 2 relative hazard [RH] = 1.25, 95% confidence interval [CI] = 0.80, 1.96; stage 3 RH = 1.76, 95% CI = 1.15, 2.70), a newly developed opportunistic disease score (Boston Opportunistic Disease Survival Score; stage 2 RH = 1.35, 95% CI = 0.90, 2.02; stage 3 RH = 2.10, 95% CI = 1.38, 3.18), and measures of activities of daily living (any intermediate limitations, RH = 1.84, 95% CI = 1.05, 3.21; any basic limitations, RH = 2.60, 95% CI = 1.44, 4.69). This model had substantially greater predictive power (R2 = .17, C statistic = .68) than the Justice score alone (R2 = .09, C statistic = .61).

CONCLUSIONS

Incorporating data on clinically important events and functional status into a physiologically based system can improve the prediction of survival with AIDS.

CT and MR imaging in the local staging of primary malignant musculoskeletal neoplasms: Report of the Radiology Diagnostic Oncology Group

PURPOSE

To assess the relative accuracies of computed tomography (CT) and magnetic resonance (MR) imaging in the local staging of primary malignant bone and soft-tissue tumors.

MATERIALS AND METHODS

At four institutions, 367 eligible patients (aged 6-89 years) with malignant bone or soft-tissue neoplasms in selected anatomic sites were enrolled. Patients underwent both CT and MR imaging within 4 weeks before surgery. In each patient, CT scans were interpreted independently by two radiologists and MR images by two other radiologists at the enrolling institution. The CT and MR images were then interpreted together by two of those radiologists and subsequently reread at the other institutions. Imaging and histopathologic findings were compared and were supplemented when needed with surgical findings. Receiver operating characteristic curve analysis and descriptive statistical analysis were performed.

RESULTS

Cases were analyzable in 316 patients: 183 had primary bone tumors; 133 had primary soft-tissue tumors. There was no statistically significant difference between CT and MR imaging in determining tumor involvement of muscle, bone, joints, or neurovascular structures. The combined interpretation of CT and MR images did not statistically significantly improve accuracy. Interreader variability was similar for both modalities.

CONCLUSION

CT and MR imaging are equally accurate in the local staging of malignant bone and soft-tissue neoplasms in the specific anatomic sites studied.

The influence of health-related quality of life and social characteristics on hospital use by patients with AIDS in the Boston Health Study

OBJECTIVES

The authors examine whether health-related quality of life (HRQL) and social factors were independent predictors of future hospital use for persons with acquired immunodeficiency syndrome (AIDS).

METHODS

A panel of 305 patients with AIDS treated at three provider settings in the Boston, Massachusetts area were enrolled during 1990 and 1991. Data were collected at baseline study enrollment and again 4 months later. Patient interviews, hospital bills, and medical charts were used to measure hospital use (admissions and days during the 4 months after enrollment), sociodemographic characteristics (age, gender, race, education, insurance, homelessness, alcohol use, and AIDS risk factors), disease burden (patient severity and a three-level opportunistic diseases and complications score), HRQL (patient-reported symptoms, activities of daily living, neuropsychological status, and global health assessment), system of care, and use of prophylactic drugs. Logistic regression was used to estimate the odds of admission. Total days of hospital care by patients with at least one admission were analyzed using multiple linear regression. Clinical models of hospital use were developed first from the variables measuring disease burden and system of care. Models estimating the associations between hospital use and all other predictor variables measured at baseline then were estimated using stepwise techniques, controlling for variables in the core model.

RESULTS

Patients were more likely than their reference groups to be hospitalized if they had serious opportunistic diseases (adjusted odds ratio [OR] = 2.7), had poorer neuropsychological status (OR = 1.9), were non-white (OR = 2.0), or were homeless (OR = 3.3) (all P < or = 0.05). Activities of daily living were associated moderately (OR = 1.3; P = 0.07). Only system of care and neuropsychological status predicted total hospital days.

CONCLUSIONS

The results indicate that future hospital use by persons with AIDS may be influenced by social and other health-related factors in addition to the more clinically related characteristics that are recorded in a medical chart. It therefore may be appropriate to assess these factors when considering options for intervention or when comparing patterns of use among patient groups or settings.

Ordinal regression methodology for ROC curves derived from correlated data

We present an approach for the analysis of correlated ROC data, using ordinal regression models in conjunction with generalized estimating equations. The approach applies to the analysis of degree-of-suspicion data derived from multiple interpretations of the same diagnostic study and from the examination of the same patients with multiple diagnostic modalities. The regression models make it possible to incorporate patient and reader characteristics into the analysis, without having to resort to stratification. We illustrate the potential of the approach with analysis of data from two studies in diagnostic oncology.

CT and MR imaging in the staging of colorectal carcinoma: report of the Radiology Diagnostic Oncology Group II

PURPOSE

To prospectively evaluate the relative accuracy of computed tomography (CT) and magnetic resonance (MR) imaging in the staging of colorectal carcinoma.

MATERIALS AND METHODS

CT and MR studies were independently interpreted in a group of 478 patients with colorectal carcinoma in a study conducted from 1989 to 1993. The accuracy of each modality was assessed in a subset of 365 patients with primary tumors with respect to staging of local extent of tumor, status of local-regional lymph nodes, and the presence of liver metastases.

RESULTS

In the staging of local extent of tumor, CT is more accurate than MR imaging, particularly in the definition of penetration of the muscularis propria by rectal cancer (74% vs 58%). Accuracies of CT and MR imaging were equivalent in depiction of transmural extent in colon cancers. CT and MR imaging exhibited accuracies of 62% and 64% in assessment of lymph node involvement with sensitivities of 48% and 22%, respectively. The accuracy of MR imaging and of CT (85% for each) are better for evaluation of liver metastases; lower sensitivities (62% and 70%, respectively) than specificities (97% and 94%, respectively) were demonstrated for both modalities.

CONCLUSION

CT was more accurate than MR imaging in detection and characterization of transmural penetration of rectal tumors. Recent technologic advances in MR imaging may affect these results.

[Management of myocardial infarction in the Rhone-Alps area. Are there many variations in practice?]

The aim of this study was to determine the diagnostic and therapeutic strategies after myocardial infarction and to examine variations in medical and surgical practice with respect to the severity of disease, status of the hospital and patients' characteristics. The method used was a prospective study with follow-up at 30 days and 18 months. The subjects came from an exhaustive cohort of all patients admitted to hospital for myocardial infarction during the month of April 1991 in 57 public and private hospitals in the Rhone-Alps region (n = 311). The patients were identified after admission by consulting physicians of the Department of Social Security. The study included all patients with acute myocardial infarction with at least two of the three usual diagnostic criteria (prolonged, constrictive chest pain, enzyme increases and electrocardiographic changes). The study excluded patients who were dead before arrival at hospital. The parameters analysed included the clinical management, use of echocardiography, exercise stress testing, myocardial scintigraphy, coronary angiography, thrombolysis, angioplasty and coronary bypass surgery in the first 30 days after admission. The severity of infarction was assessed by seven clinical, enzymatic and electrocardiographic criteria by physicians from the Department of Social Security (pain, syncope, shock, left ventricular dysfunction, elevation of CPK > 1000 IU, anterior or extensive necrosis, arrhythmias). The mortality rate of this cohort was calculated from hospital statistics and then by enquiring in the town halls of the region. The demographic features of the cohort were marked by a predominance of men (69.5%) and a relatively high mean age (69 years; 23% over 80 years). Complementary investigations were used with the following frequencies: echocardiography, 61.1%; coronary angiography, 26.4%; exercise stress testing 22.8%; myocardial scintigraphy, 5.5%.(ABSTRACT TRUNCATED AT 250 WORDS)