Clinical utility and management impact of M-mode echocardiography

Transthoracic echocardiography in pediatric intensive care: impact on medical and surgical management

OBJECTIVES

Although transthoracic echocardiography is commonly performed in the PICU, its utility is not specifically known. The purpose of this investigation was to evaluate the clinical impact of echocardiography in the PICU in terms of frequency of unanticipated findings and the frequency and nature of clinical management changes attributed to the results of echocardiography.

DESIGN

Prospective cohort study.

SETTING

Nineteen-bed combined medical-surgical-cardiac PICU at a tertiary care children's hospital.

PATIENTS

All patients in PICU undergoing transthoracic echocardiography.

INTERVENTIONS, MEASUREMENTS, AND MAIN RESULTS

Data collected included echocardiography indications, pre-echocardiography clinical assessment of anticipated echocardiography findings, height, weight, primary diagnosis, age, and urgency (stat vs routine) of echocardiography. Input of the attending care team (intensivist, cardiologist, and/or cardiovascular surgeon) allowed classification of echocardiography results as either confirming the pre-echocardiography impression, altering the pre-echocardiography clinical impression regarding the indication for which the test was performed, or altering the impression by virtue of new findings unrelated to the specific indication. The nature of the new findings were recorded and categorized. The team recorded clinical management changes made in response to the echocardiography results; the nature of these were listed and categorized. Echocardiograms (n = 416) were performed in 132 patients. Of these, 244 echocardiograms (59%) were ordered on male patients, 31% were under 30 days old, median age was 103 days, 379 (91%) had a primary cardiac diagnosis, and 92 (22%) were ordered stat. Sixty-three percent of echocardiograms confirmed and 24% altered the pre-echocardiography impression regarding the indication for the echocardiography; 13% introduced new findings unrelated to the indication. Cardiac surgical revision was the management change required in 26 patients (6.3%). Stat echocardiography was more likely to alter the pre-echocardiography assessment than routine echocardiography (p < 0.001). Management changes were more commonly associated with stat echocardiograms (p = 0.002) and those with new unexpected findings (p < 0.001) but had no demonstrable association with age less than 30 days (p = 0.332).

CONCLUSIONS

Unanticipated echocardiography results are common in the PICU, and they often alter the clinical impressions that prompted the echocardiogram or introduce new findings unrelated to the reason for which the echocardiogram was recorded. Clinical management changes attributable to echocardiography findings are frequent in the PICU, including occasional surgical intervention. Echocardiography adds diagnostic value and contributes to the management approach in the PICU, accounting for its frequent use.

Screening cardiac ultrasonographic examination in patients with suspected cardiac disease in the emergency department

BACKGROUND/OBJECTIVE

Our purpose was to evaluate the utility of a brief screening cardiac ultrasonographic (SCU) examination. We prospectively compared the SCU with conventional clinical evaluation in 124 emergency department (ED) patients with suspected cardiac disease. Furthermore, we assessed the impact and quality of SCU examinations as obtained by briefly trained ED personnel (EP).

METHODS

Patients underwent clinical evaluation by an ED physician and SCU examination by a sonographer or cardiologist. Patient disposition, hospital stay length, and the number of full echo examinations were compared with the presence of significant findings on SCU. In patients who received a full echocardiogram during hospitalization, results of the initial clinical examination were compared with results of the SCU examination in the diagnosis of significant findings. A similar analysis, but with quality assessment, was performed on only those SCU examinations acquired by 4 EP.

RESULTS

Of the 124 patients enrolled in the main study, 40 of 124 (32%) had significant findings on SCU. Of patients with abnormal SCUs versus normal SCUs, 16 of 40 (40%) versus 18 of 84 (21%) had hospital stay lengths >2 days (P < or =.05). Using the 36 inpatient full echo studies obtained for standard indications during hospitalization as a gold standard, initial clinical examination identified only 7 of 30 (23%) significant findings and had 16 false-positive diagnoses, whereas SCU identified 22 of 30 (73%) with 8 false positives. Although similar study results occurred with interpretation of 68 SCUs obtained by EP, quality was achieved in only 55% ED personnel versus 97% of sonographer-obtained SCUs (P <.05).

CONCLUSIONS

An SCU examination detects significant findings misdiagnosed on initial clinical evaluation in the ED and provides prognostic data regarding length of hospital stay.

Ordering an echocardiogram for evaluation of left ventricular function: level of expertise necessary for efficient use

This prospective study was performed to test the hypothesis that the yield of 2-dimensional echocardiography (2DE) would be higher when it is ordered by a cardiologist than by a noncardiologist. Patients referred for transthoracic 2DE for the evaluation of left ventricular systolic function for the 11-month period between July 10, 1995, and June 10, 1996, were included in the study. Demographic, historical, and clinical findings were recorded. Whether the patient was referred by a cardiologist versus a noncardiologist was used as the predictor variable in a binary logistic regression analysis. To address the possibility that the yield of 2DE may be higher for cardiologists because the prevalence of disease in patients referred to them may be higher (selection bias), the analysis was subjected to a propensity score adjustment. Of 2176 patients referred for 2DE during the study, 1033 were referred for the evaluation of left ventricular function. The test had a positive yield in 52% of patients for cardiologists versus 31% for noncardiologists (chi(2) = 45.5, P <.0001, odds ratio 2.4 [CI = 1. 9-3.1]). This difference remained highly significant even when propensity score risk adjustment was made (chi(2) = 54.2, P <.0001, odds ratio 2.0 [CI = 1.5-2.8]). We conclude that the yield of 2DE is higher for cardiologists compared with noncardiologists and that this result was not related to differences in patient populations examined by the two groups. Thus, more efficient use of 2DE may be achieved if patients are referred to cardiologists rather than directly sent for 2DE.

What is the real clinical utility of echocardiography? A prospective observational study

This was a prospective, structured interview to evaluate physician expectations of echocardiography and the subsequent impact on patient care. The setting involved requests for echocardiograms in patients admitted to a tertiary care teaching hospital. Measurements included assessment of the diagnostic and therapeutic implications of echocardiography perceived by the physicians and subsequently reported and confirmed by blinded chart review. From January to May 1997, 542 echocardiograms were performed on 500 inpatients (age 62 +/- 17 years; 56% men). Referring physicians were mainly house staff (83%) and from medical services. The main indications were evaluation of left ventricular function (54%) or valve function (16%). At the time of the request, 89% of physicians believed that echocardiography was needed to guide future investigations or treatment, although in 24% of cases, physicians could not provide details of such. A reported change in treatment occurred in 57% and was validated by chart review in 38%. Changes occurred more frequently in patients in the intensive care unit versus those not in the intensive care unit (54% vs 37%, P =.02) but were similar between attending physicians and house staff (frequency of change 41% vs 39%, P = not significant) and between those with and those without previous echocardiograms (38% vs 39%, P = not significant). The utility of the echocardiogram to influence treatment decisions in hospitalized patients is high, especially in critically ill patients. However, this impact is less than is anticipated at the time of the initial request. Further studies involving more select populations of patients are required.

Clinical usefulness and cost of echocardiography in patients admitted to a coronary care unit

The usefulness and cost of echocardiography was evaluated in 133 consecutive patients admitted to the Coronary Care Unit. A useful echocardiogram was one that provided new information, which influenced diagnosis, prognosis, or treatment. The cost of a useful echocardiogram was defined as the unit cost ($476 the Medicare global fee) x units (i.e., total echocardiograms / useful echocardiograms). Admission diagnoses were unstable angina (34%), arrhythmia (14%), congestive heart failure (8%), postprocedure monitoring (7%), acute myocardial infarction (6%), and miscellaneous (20%). The echocardiogram provided new information in 29% of patients. Patients without a recent echocardiogram (within 3 months) were twice as likely to have a useful echocardiogram (33 of 99, 33%) as those with a recent echocardiogram (5 of 34, 15%, p <0.05). A cardiologist predicted the overall usefulness of echocardiography with a positive predictive accuracy of 52% and a negative predictive accuracy of 94% (p < 0.0001). The overall cost of a useful echocardiogram of 3.5 units or $1,666 per useful study was decreased to $904 (1.9 units) if only studies predicted to be useful were considered. The usefulness of echocardiography varied significantly (p <0.02) within the admitting diagnostic categories. The usefulness of an echocardiogram was underestimated in patients with congestive heart failure, where it was found to be most useful (64%; $762 or 1.6 units). Thus, usefulness relates to the admission diagnosis, the availability of a recent echocardiogram, and to clinical judgment.

The clinical impact of echocardiography on antibiotic prophylaxis use in patients with suspected mitral valve prolapse

PURPOSE

To determine the impact of echocardiography on the use of antibiotic prophylaxis in patients with suspected mitral valve prolapse (MVP).

PATIENTS AND METHODS

We evaluated 147 consecutive patients who were referred for "rule out mitral valve prolapse" to a university hospital echocardiography laboratory. Chart review and phone contact were used to determine the demographic characteristics of the patients; past diagnosis of MVP, symptoms, and exam at referral; practice specialty of referring MD; echocardiographic findings; and change in prophylaxis usage as a result of the echocardiogram (ECHO). Prophylaxis was considered to be indicated if the echocardiogram demonstrated MVP with at least mild regurgitation or abnormal thickening of at least one mitral leaflet.

RESULTS

Based on the ECHO a change in antibiotic prophylaxis was indicated in 20 of 147 (14%) patients including initiation of prophylaxis in 6, and discontinuation of prophylaxis in 14. However, only 4 of 20 patients (20%) actually changed their prophylaxis habits leading to an actual yield of 4 management changes per 131 ECHOs ordered (3%). This corresponded to 1 change in management per $36,250 in hospital and physician costs. Younger age, female gender, and presence of symptoms were associated with a benign ECHO. Indications for a change in management were not significantly different between physician specialities: 18% for generalists (internal medicine and family practice), 12% for cardiologists, and 7% for other specialists, P = 0.3.

CONCLUSIONS

In patients referred for evaluation of MVP, echocardiography infrequently resulted in changes in antibiotic prophylaxis management and was associated with significant expense.

The use and interpretation of commercial APC gene testing for familial adenomatous polyposis

BACKGROUND

The use of commercially available tests for genes linked to familial cancer has aroused concern about the impact of these tests on patients. Familial adenomatous polyposis is an autosomal dominant disease caused by a germ-line mutation of the adenomatous polyposis coli (APC) gene that causes colorectal cancer if prophylactic colectomy is not performed. We evaluated the clinical use of commercial APC gene testing.

METHODS

We assessed indications for APC gene testing, whether informed consent was obtained and genetic counseling was offered before testing, and the interpretation of the results through telephone interviews with physicians and genetic counselors in a nationwide sample of 177 patients from 125 families who underwent testing during 1995.

RESULTS

Of the 177 patients tested, 83.0 percent had clinical features of familial adenomatous polyposis or were at risk for the disease-both valid indications for being tested. The appropriate strategy for presymptomatic testing was used in 79.4 percent (50 of 63 patients). Only 18.6 percent (33 of 177) received genetic counseling before the test, and only 16.9 percent (28 of 166) provided written informed consent. In 31.6 percent of the cases the physicians misinterpreted the test results. Among the patients with unconventional indications for testing, the rate of positive results was only 2.3 percent (1 of 44).

CONCLUSIONS

Patients who underwent genetic tests for familial adenomatous polyposis often received inadequate counseling and would have been given incorrectly interpreted results. Physicians should be prepared to offer genetic counseling if they order genetic tests.

Utilization of echocardiography by internists and cardiologists: a comparative study

OBJECTIVE

To compare the utilization of echocardiography as a diagnostic tool by internists and cardiologists.

DESIGN

Retrospective study.

SETTING

Tertiary care university hospital.

METHODS

Indications and clinical utility of echocardiographic studies ordered by cardiologists (group A, n = 301) and internists (group B, n = 297) were compared by chart review. The two groups of patients were analyzed to determine if the studies detected new cardiac pathology and/or altered patient management.

RESULTS

The proportion of studies with abnormal results were similar in both groups (19% versus 14%, P > 0.05). The results of echocardiography, however, led to a change in management more often when the study was ordered by cardiologists (16% versus 10%, P < 0.05). A significantly greater proportion of studies were ordered for evaluation of valvular function by internists (44% versus 33%, P < 0.05). Echocardiography detected valvular abnormalities in a similar proportion of cases in groups A and B (14% versus 10%, P > 0.05). However, diagnostic yield was very poor when the study was performed in patients with suspected mitral valve prolapse in both groups. Cardiologists utilized echocardiography more often for evaluation of left ventricular function (35% versus 18%, P < 0.01) and in the setting of atherosclerotic heart disease for detecting wall motion abnormalities (14% versus 5%, P < 0.01).

CONCLUSION

Diagnostic yield of echocardiography is similar when ordered by internists and cardiologists. With the information obtained, management is altered in a slightly greater proportion of cases involving a cardiologist. This may be due to utilization of echocardiography more often for estimating left ventricular function and for detecting wall motion abnormalities by cardiologists.

Where Does Ultrasound Stand in Technology Assessment and Outcomes Research

Although diagnostic ultrasound has been widely applied for various clinical conditions and diseases, its clinical effectiveness remains less well defined. Technology assessment can be applied to ultrasound and includes several criteria such as techologic capability, range of possible uses, diagnostic accuracy, impact on health care providers, therapeutic impact, and patient outcome. A review of each of these criteria is provided to raise sonographer awareness as to the need for further reserch in this area.

The appropriateness of use and the clinical impact of micturating cystourethrography in paediatric practice

STUDY OBJECTIVE

To determine the appropriateness of use and the clinical impact of micturating cystourethrography (MCU) in paediatric practice.

DESIGN

Retrospective medical record review.

SETTING

A major teaching children's hospital in the Trent region.

PATIENTS

Consecutive sample of 120 children undergoing MCU during 1991-1992, identified from the radiology records.

MAIN OUTCOME MEASURES

The referring clinician's reasons for requesting an MCU; the clinical management plan pursued before and after the MCU, and the change in management initiated by the result of the examination; the appropriateness of use of the test, as determined by the presence or absence of a valid indication for MCU in the specific clinical situation.

RESULTS

The change in patient management attributable to the MCU result could be categorized as: no change (19%); decision to end the imaging investigation of the patient (33%); decision to end all investigations, and prophylactic or therapeutic interventions (16%); decision to discontinue chemoprophylaxis (2%); decision to end the imaging investigation and introduce chemoprophylaxis and follow-up for bacteriuria (6%); decision to continue the imaging investigation and introduce chemoprophylaxis and follow-up for bacteriuria (13%); and decision to operate or help in planning the surgical treatment required (11%). Inappropriate use of the test was observed in 20% of the cases.

CONCLUSIONS

This study provides a basis for understanding the use of MCU in paediatric practice. The findings that 19% of the cystourethrograms had no appreciable effect, and that 20% of the examinations were used inappropriately show the need for increased effort to minimize overuse of the test.

Clinical utility of transthoracic two-dimensional and Doppler echocardiography

OBJECTIVES

The purpose of this study was to determine the value of contemporary echocardiography for patient diagnosis and management in clinical practice.

BACKGROUND

Although the use of echocardiography is growing rapidly, there is little information about its clinical utility.

METHODS

A prospective observational study was performed at a community-based, tertiary care teaching hospital. Physicians were interviewed before and after learning the result of their patients' echocardiographic examination. Chart reviews were performed to confirm reports of new diagnoses and treatments that resulted from echocardiography.

RESULTS

Physician interviews were successfully completed for 244 (49%) of 497 inpatients and 101 (30%) of 336 outpatients. Among patients with moderate or greater mitral regurgitation, the diagnosis of "clinically significant" mitral regurgitation was unsuspected in 5 (31%) of 16 outpatients and 28 (60%) of 47 inpatients. Among the patients with left ventricular wall motion abnormalities, the result was unsuspected in 7 (50%) of 14 outpatients and 18 (22%) of 81 inpatients. As assessed by chart review, echocardiography produced a definite new diagnosis in 25 inpatients (10%) and 5 outpatients (5%) and was responsible for changes in pharmacologic treatment in at least 16 inpatients (7%) and 2 outpatients (2%).

CONCLUSIONS

The echocardiogram commonly provided information that was unexpected. This information provided a definite new diagnosis or treatment in a smaller proportion of the patients. Further research is necessary to define the appropriate yield that would warrant echocardiography on the basis of both clinical and cost-effectiveness criteria.

Age-related variability in the use of cardiovascular imaging procedures

OBJECTIVE

To determine if older patients undergo fewer cardiovascular imaging procedures (CIPs) than younger patients when admitted to a tertiary care academic medical center for an acute myocardial infarction (MI), after adjusting for disease severity and comorbidities.

DESIGN

Non-current prospective cohort study.

SETTING

Urban tertiary care academic medical center.

PATIENTS

Medical records of 294 patients admitted and diagnosed with an acute MI between January 1990 and April 1991 were reviewed.

MEASUREMENTS

The total number of different CIPs performed during hospitalization was determined. Cardiac catheterizations, echocardiograms, radionuclide ventriculograms, and thallium scans counted as CIPs. Disease severity was assessed by the Acute Physiology Score (APS) of APACHE II, admission Killip's Classification, and peak creatine phosphokinase (CPK) levels. Comorbidities were assessed using a modified Comorbidity Damage Index of Charlson.

RESULTS

The mean (+/- SD) number of different CIPs performed during hospitalization was significantly less for those > or = 75 years old (1.3 +/- 1.0) than for those < 75 years old (1.7 +/- 1.0) (P = 0.01), and CIP number negatively correlated with age (Spearman r = -0.178; P = 0.01). Mean CIP number decreased from 2.0 +/- 1.1 for those < 45 years old to 0.9 +/- 0.6 for those > or = 85 years old (P = 0.02). Other factors positively associated (P < 0.10) with CIP number were: CPK values in the highest quartile of the study population (> 355 U/L); admission to a cardiology, medical, or family practice service; no CIP performed at an outside hospital prior to transfer; admission Killip's Classification of less than IV, and a Q-wave MI. After adjusting for these variables in a multiple regression model, age > or = 75 remained an independent predictor of decreased CIP use (P = 0.003). The modified comorbidity index score and the APS score, a general measure of severity of illness, were not significantly associated with CIP use. When procedures were examined individually, no significant age-related differences were noted in the use of thallium scans, radionuclide ventriculograms, or echocardiograms. Older patients did, however, remain less likely to undergo cardiac catheterizations (P < 0.001).

CONCLUSION

Older patients, regardless of underlying disease severity or comorbidities, undergo fewer invasive cardiovascular evaluations than younger patients when admitted to a tertiary care academic medical center for an acute MI.

The clinical utility of transcranial Doppler ultrasound in suspected vertebrobasilar ischemia

The impact of transcranial Doppler ultrasound (TCD) on the management of 107 consecutively referred patients with suspected vertebrobasilar ischemia was studied. Physicians who referred patients for TCD of the posterior circulation were interviewed before and after being given the results of the TCD examination. Management plans devised prior to knowledge of the TCD results were compared to those devised after TCD results were given. Management changed in 42% of the patients after TCD results were given, and there was a 58% reduction in the use of angiography (p = 0.04) and a 128% increase in the use of aspirin (p = 0.005). Changes in the use of anticoagulants and other diagnostic or therapeutic approaches were not significant. TCD appears to have a significant impact on the management of patients with suspected vertebrobasilar ischemia.

Utility of echocardiography in patients with suspected mitral valve prolapse

PURPOSE

Echocardiography has become a widely utilized test since its introduction into clinical medicine in the early 1970s. Although it has frequently been performed in patients suspected of having mitral valve prolapse (MVP), its usefulness in this setting has not been systematically studied. To investigate the use and value of echocardiography in patients suspected of having MVP, we conducted a prospective study in which physicians were interviewed before and after ordering echocardiographic testing for patients in whom there was a suspicion of MVP.

PATIENTS AND METHODS

The study population included consecutive patients referred to the echocardiography laboratory at Boston University Medical Center because of suspected MVP between January 1 and December 31, 1987. Two standardized telephone interviews were conducted with the physician most responsible for ordering the echocardiogram. The following information was obtained during the first interview, which was always conducted before the echocardiogram was performed: patient demographic and clinical data; the reason for ordering the echocardiogram; the physician's most likely clinical diagnosis; the physician's estimate of the likelihood that the patient had MVP; and the physician's proposed management plans. After the referring physician received the echocardiographic results, a second interview was conducted to determine changes in the most likely clinical diagnosis and management plans. The impact of the echocardiogram on diagnosis and management was evaluated by comparing physician responses before and after reception of echocardiographic results. Receiver operating characteristic (ROC) curves were constructed to assess the physician's skills at distinguishing patients with echocardiographic-documented MVP from those without MVP.

RESULTS

A total of 106 echocardiograms were ordered by 45 different physicians. More than 80% of all echocardiograms were ordered to address diagnostic or therapeutic concerns. On echocardiography, 47 (44%) patients were found to have MVP, six (6%) had mitral regurgitation without prolapse, and 53 (50%) had normal results. On the basis of the ROC curve analysis, the physician's ability to discriminate between patients with and without echocardiographic MVP varied significantly by physician specialty and practice setting. The echocardiographic results led to a change in diagnosis in 59 (56%) patients. A change in management occurred in 29 (27%) patients, with 25 of these 29 changes (86%) related to the initiation or discontinuation of antibiotics.

CONCLUSIONS

Echocardiography frequently alters diagnostic assessments and leads to therapeutic changes in some patients suspected of having MVP. However, the benefits of such changes have not yet been demonstrated.

The contribution of a non-invasive test to clinical care. The impact of echocardiography on diagnosis, management and patient anxiety

We prospectively studied the impact of echocardiography on a cardiologist's diagnosis and management plan and on patient anxiety for 300 consecutive referrals. There was an impact on diagnosis in 90% of patients. Most common was confirmation of diagnosis usually with the addition of information pertinent to management (81%); change of disease category or resolution of diagnostic doubt was uncommon (9%). The consultant cardiologist believed the heart to be normal in 48 patients who did not have any associated disease; none had any echocardiographic abnormality. The cardiologist reported increased diagnostic confidence in 74% of all patients but management changed in only 9%. One-third of all patients reported reduced anxiety when this was an important clinical issue but in less than half of them did the cardiologist consider that echocardiographic information was essential for reassurance. Anxiety was increased in 6%, and in 12% the anxiety response was inconsistent with the test result. The clinical contribution of the test report was most obvious for those patients (30%) in whom this data was required for (a) a decision concerning specific diagnostic or technical intervention (b) a change of management plan which implied obvious or likely health benefit (c) reassurance which was a clinically important issue. The magnitude of this contribution was related to the study indication. The value of echocardiography is obvious when assessing patients for invasive intervention or when proper treatment or adequate reassurance are impeded by diagnostic doubt. However, for many current indications, we need better definition of factors which predict a clinically useful result. In particular, when the aim is to rule-out disease, our results suggest that an expert cardiological opinion would often be more appropriate than an echocardiogram.

Exercise thallium scans: patterns of use and impact on management of patients with known or suspected coronary artery disease

Physicians' ordering patterns and subjective assessments of the management impact of 345 exercise thallium studies at a tertiary care hospital were evaluated. Exercise thallium tests were ordered to obtain functional information (60%), to detect myocardial ischemia (29%), for routine follow-up (6%) and for screening for coronary artery disease (CAD) (5%). Exercise thallium testing was ordered instead of exercise electrocardiography alone because of an interest in assessing the size of an ischemic or infarcted area (37%), a belief that the exercise electrocardiogram was likely to be nondiagnostic (35%) or that exercise thallium was a better test (23%). In 20% of cases physicians said that thallium imaging results were helpful or pivotal in major management decisions, such as those regarding catheterization or surgery. In most cases, no further tests were ordered and antianginal therapy or activity prescription was changed after exercise thallium testing. Exercise thallium images also helped to stratify patients into low and high probability of CAD categories better than exercise electrocardiographic results alone did. Our data suggest that physicians are using exercise thallium testing to obtain functional information more often than simply to detect CAD and that physicians consider the test to be of considerable value.

Guidelines for the clinical and economic evaluation of health care technologies

The health care system is routinely confronted with promising new technologies. In the past, most new technologies have been integrated into clinical practice without a rigorous demonstration of their effectiveness or efficiency. In order to provide a more rational approach to the adoption and utilization of health technology a comprehensive set of guidelines for both clinical and economic evaluation is proposed. While conceived of as an ideal that is unlikely to be universally met in practice, it is argued that decision making can be improved by striving towards this goal. The clinical guidelines stress the advantages of subjecting major new technologies to randomized controlled trials and insisting upon a demonstration of patient benefit in the application of diagnostic technologies. The economic guidelines stress comparisons with relevant alternative uses of the resources and the assessment of the impact on the quality of life. While application of the guidelines will produce rigorous and useful evidence, the final decisions concerning the allocation of health care resources must rest fundamentally on social value judgements and not solely, or even primarily, on informed expert opinion.

The role of before-after studies of therapeutic impact in the evaluation of diagnostic technologies

Studies in which clinicians' therapeutic plans are determined before and after application of a new test offer an attractive alternative to randomized trials as a strategy for establishing the usefulness of diagnostic technologies. To maximize scientific credibility such studies should include a number of design features: specific clinical questions should be addressed; consecutive patients presenting with a clinical problem should be entered; the test's accuracy should be assessed; and changes in therapy should be described in sufficient detail to allow an estimate of the impact on patient outcome. Even with attention to these issues, the validity of before-after studies of therapeutic impact is limited by possible discrepancies between clinicians' stated plans and actual behavior, and by the often tenuous link between change in therapy and patient benefit. Before-after designs will be most useful in primary investigations which, if positive, should be followed by more rigorous studies to establish patient benefit.