Risk of Upper Gastrointestinal Injury and Events in Patients Treated With Cyclooxygenase (COX)-1/COX-2 Nonsteroidal Antiinflammatory Drugs (NSAIDs), COX-2 Selective NSAIDs, and Gastroprotective Cotherapy

Numerous studies using varying methodologies and outcome measures have examined the gastrointestinal risks of aspirin and nonaspirin nonsteroidal antiinflammatory drug (NSAID) use. Despite the large volume of literature, clarity regarding the key risk factors and their quantitative importance is lacking. We performed a comprehensive review of the literature to summarize the incidence of gastrointestinal injury in populations with varying risk characteristics using agents that inhibit both isoforms of cyclooxygenase and those that selectively inhibit only cyclooxygenase-2 (COX-2).Although risk estimates vary, the risk of serious gastrointestinal complications in NSAID users is approximately 2.5 to 4.5 times that of nonusers. The risk of NSAID-related gastrointestinal bleeding is augmented by concomitant low-dose aspirin and could approach double the risk of NSAID use alone. The preponderance of evidence shows that the risk of NSAID-related gastrointestinal bleeding is reduced approximately 50% with a coxib as compared with traditional NSAID. The relative risk of hospitalization resulting from upper gastrointestinal bleeding for patients treated with a nonselective NSAID was 4.4 (95% confidence interval [CI], 2.3-8.5) and 1.9 (95% CI, 1.0-3.5) when compared with celecoxib and rofecoxib, respectively. Aspirin increases the risk of NSAID-related gastrointestinal bleeding in patients taking COX-2 selective inhibitors, with odds ratios ranging from 5.8 to 7.7; however, it is unknown whether this risk is greater than the risk from aspirin alone. The risks from both traditional NSAIDs and COX-2 inhibitors are increased in the elderly, patients on anticoagulation, and patients with prior gastrointestinal events.Gastroprotective agents have been found to significantly reduce the risk for gastrointestinal injury in patients receiving NSAID therapy, especially those receiving concurrent low-dose cardioprotective doses of aspirin. Proton pump inhibitors (PPIs) and misoprostol both reduce the incidence of gastric and duodenal ulcers, as well as recurrence of ulcer complications in patients receiving NSAIDs. The relative risk for gastric ulcers ranged from 0.17 to 0.38, whereas for duodenal ulcers, the range was 0.11 to 0.28. Although misoprostol is slightly more effective in preventing gastric ulcers in these patients, PPIs are better tolerated. Although NSAIDs appear safe in "low-risk" populations, our review suggests that the use of gastroprotective cotherapy should be considered in patients at higher risk of NSAID-related upper gastrointestinal bleeding.

Pharmaceutical promotion: don't throw the baby out with the bathwater

Spending on prescription drugs and promotion by the pharmaceutical industry grew substantially during the past ten years. Does the greater exposure offered by promotion fill a needed educational gap, or does it merely promote inappropriate use? This paper uses two recent studies to explore this question, presenting a framework in which the impact of promotion depends upon the level of evidence and consensus on drug use.

Guidelines for the appropriate use of non-steroidal anti-inflammatory drugs, cyclo-oxygenase-2-specific inhibitors and proton pump inhibitors in patients requiring chronic anti-inflammatory therapy

AIM

To rationalize decision making around the use of different non-steroidal anti-inflammatory drug (NSAID) treatment strategies in patients with varying degrees of gastrointestinal and cardiovascular risk.

METHODS

The panel comprised nine physicians (three rheumatologists, two internists, two gastroenterologists and two cardiologists) from geographically diverse areas practising in community-based settings (n = 4) and academic institutions (n = 5). A literature review was performed by the authors on the risks, benefits and costs of NSAIDs, cyclo-oxygenase-2-specific inhibitors and proton pump inhibitor co-therapy. The RAND/UCLA Appropriateness Method was used to rate 304 clinical scenarios as 'appropriate', 'uncertain' or 'inappropriate'.

RESULTS

In patients with no previous gastrointestinal event and not concurrently on aspirin (low risk), the panel rated the use of an NSAID alone as 'appropriate' for those aged < 65 years, and the use of an NSAID +proton pump inhibitor or cyclo-oxygenase-2-specific inhibitor + proton pump inhibitor as 'inappropriate'. For patients aged > 65 years and at low risk, an NSAID or cyclo-oxygenase-2-specific inhibitor alone was rated as 'uncertain'. For patients with a previous gastrointestinal event or who concurrently received aspirin, an NSAID alone was rated as 'inappropriate', and either a cyclo-oxygenase-2-specific inhibitor or an NSAID +proton pump inhibitor was rated as 'appropriate'. Finally, for patients with a previous gastrointestinal event and on aspirin, an NSAID or cyclo-oxygenase-2-specific inhibitor in conjunction with a proton pump inhibitor was rated as 'appropriate'.

CONCLUSIONS

Clinicians and managed care entities need to balance the risks, benefits and costs of NSAIDs, cyclo-oxygenase-2-specific inhibitors and the prophylactic use of proton pump inhibitors. The guidelines given here can assist this process.

Growth in use of lipid-lowering therapies: are we targeting the right patients?

BACKGROUND

Prescription drug spending has been rising at > 10% per year, with volume of use (rather than price) being the primary driver for that growth. Concern exists that industry marketing has led to increased use of medications by patients with marginal indications.

OBJECTIVE

To determine whether the increase in the number of patients receiving lipid-lowering therapy represents a shift away from treatment of patients at highest cardiovascular (CV) risk towards patients in lower risk categories.

STUDY DESIGN AND METHODS

Cardiovascular risk criteria adapted from guidelines were applied to an administrative database of medical and pharmaceutical claims for 1997 and 1999 that included managed care plan enrollees in 22 states. Patients were assigned to 1 of 7 categories representing CV risk based on documentation of CV disease/risk factors, with category 1 and 2 indicating the highest risk group (secondary prevention). The odds of the treated population being in the highest risk during 1997 versus 1999 were calculated, adjusting for age and sex.

RESULTS

Patients treated with lipid medications in the study population increased from 5% in 1997 to 8% in 1999. During the same period, the percentage of treated patients in categories 1 through 6 rose from 17% to 21%. The odds of the treated population being in the highest risk group did not differ significantly between the 2 years (odds ratio (OR) = 0.99; 95% confidence interval, (CI) 0.96-1.01; P = .40).

CONCLUSIONS

Despite an increase in the percentage of patients receiving lipid-lowering therapy from 1997 to 1999, treatment rates rose modestly across all categories. Greater overall use did not appear to be associated with a shift in use towards patients with less CV risk.

Outpatient cancer drug costs: changes, drivers, and the future

BACKGROUND

To the authors' knowledge, no analysis has examined the specific components of drug spending for overall cancer care. The authors' objective was to quantify and characterize trends in outpatient drug expenditures for cancer patients.

METHODS

The authors retrospectively analyzed pharmacy and outpatient professional claims data from commercial and Medicare health maintenance organization enrollees with a solid tumor diagnosis in 1995 and 1998. Charges were subdivided by type of drug (antineoplastic drugs, chemotherapy adjuncts, supportive drugs, and drugs unrelated to cancer treatment).

RESULTS

In 1995, 14,663 cancer patients received outpatient drug treatment and 13,829 patients in 1998. Total charges increased from $17.9 million (mean charge of $1218 per patient) to $27.9 million (mean charge of $2003 per patient), an average annual increase of 16%. Antineoplastic therapy constituted the largest component of cancer-related drug costs (67%) and represented 76% of the increase from 1995 to 1998. Most charges were incurred in the professional setting for agents administered by injection. The primary explanation for the increases appeared to be a shift in treatment patterns toward newer, more expensive antineoplastic agents. Supportive therapy represented 17% of the increase in cancer drug costs, followed by chemotherapy adjuncts (7%). Charges for drugs unrelated to cancer therapy increased by 21% per year.

CONCLUSIONS

Antineoplastic therapy administered in an office or clinic was the single most important cost driver, with newer more expensive agents replacing older, less expensive drugs. Attempts to understand and control outpatient drug cost increases for cancer patients should focus primarily on antineoplastic therapy, especially the appropriate substitution of newer agents for older, less expensive alternatives. Some non-chemotherapy cancer drugs may offer an opportunity to improve quality of life with a relatively small effect on overall cancer drug costs.

Geographic variation in the use of medications: is uniformity good news or bad?

Studies have repeatedly found much geographic variation in use of surgical and diagnostic procedures. This study of the variability of medication use for specific conditions in eleven California regions finds surprisingly few differences among regions. The difference between the highest- and lowest-use areas was far less than we anticipated and amounted to only 30-40 percent for many drugs. We explore five potential explanations for low geographic variability: financial incentives, impact of managed care, study design elements, characteristics of California, and pharmaceutical marketing and education efforts. To determine whether these findings represent good or bad news will require further study.

Developing indications for the use of sentinel lymph node biopsy and adjuvant high-dose interferon alfa-2b in melanoma

OBJECTIVES

To convene a multidisciplinary panel of dermatologists, surgical oncologists, and medical oncologists to formally review available data on the sentinel lymph node (SLN) biopsy procedure and high-dose adjuvant interferon alfa-2b therapy for patients with melanoma and to rate the "appropriateness," "inappropriateness," or "uncertainty" of the procedure and therapy to guide clinical decision making in practice.

PARTICIPANTS

The panel comprised 13 specialists (4 dermatologists, 4 oncologists, and 5 surgeons) from geographically diverse areas who practiced in community-based settings (n = 8) and academic institutions (n = 5). Participants were chosen based on recommendations from the relevant specialty organizations.

EVIDENCE

A formal literature review was conducted by investigators at Protocare Sciences Inc, Santa Monica, Calif, on the risks and benefits of performing an SLN biopsy in patients with stage I or II melanoma and adjuvant interferon alfa-2b therapy in patients with stage II or III disease. The MEDLINE database was searched from 1966 through July 2000, and supplemental information was obtained from various cancer societies and cancer research groups. Panel participants were queried on additional sources of relevant information. Unpublished, presented data were included in abstract form on 1 recently closed clinical trial.

CONSENSUS PROCESS

The RAND/UCLA Appropriateness Method was used to review and rate multiple clinical scenarios for the use of SLN biopsy and interferon alfa-2b therapy. The consensus method did not force agreement.

CONCLUSIONS

The panel rated 104 clinical scenarios and concluded that the SLN biopsy procedure was appropriate for primary melanomas deeper than 1.0 mm and for tumors 1 mm or less when histologic ulceration was present and/or classified as Clark level 4 or higher. The SLN biopsy was deemed inappropriate for nonulcerated Clark level 2 or 3 melanomas 0.75 mm or less in depth and uncertain in tumors 0.76 to 1.0 mm deep unless they were ulcerated or Clark level 4 or higher. Interferon alfa-2b therapy was deemed appropriate for patients with regional nodal and/or in-transit metastasis and for node-negative patients with primary melanomas deeper than 4 mm. The panel considered the use of interferon alfa-2b therapy uncertain in patients with ulcerated intermediate primary tumors (2.01-4.0 mm in depth) and inappropriate for node-negative patients with nonulcerated tumors less than 4.0 mm deep. Specialty-specific ratings were conducted as well.

Overuse and underuse of colonoscopy in a European primary care setting

BACKGROUND

Efforts to decrease overuse of health care may result in underuse. Overuse and underuse of colonoscopy have never been simultaneously evaluated in the same patient population.

METHODS

In this prospective observational study, the appropriateness and necessity of referral for colonoscopy were evaluated by using explicit criteria developed by a standardized expert panel method. Inappropriate referrals constituted overuse. Patients with necessary colonoscopy indications who were not referred constituted underuse. Consecutive ambulatory patients with lower gastrointestinal (GI) symptoms from 22 general practices in Switzerland, a country with ready access to colonoscopy, were enrolled during a 4-week period. Follow-up data were obtained at 3 months for patients who did not undergo a necessary colonoscopy.

RESULTS

Eight thousand seven hundred sixty patient visits were screened for inclusion; 651 patients (7.4%) had lower GI symptoms (mean age 56.4 years, 68% women). Of these, 78 (12%) were referred for colonoscopy. Indications for colonoscopy in 11 patients (14% of colonoscopy referrals or 1.7% of all patients with lower GI symptoms) were judged inappropriate. Among 573 patients not referred for the procedure, underuse ranged between 11% and 28% of all patients with lower GI symptoms, depending on the criteria used.

CONCLUSIONS

Applying criteria from an expert panel of nationally recognized experts indicates that underuse of referral for colonoscopy exceeds overuse in primary care in Switzerland. To improve quality of care, both overuse and underuse of important procedures must be addressed.

Is the diagnostic yield of upper GI endoscopy improved by the use of explicit panel-based appropriateness criteria?

BACKGROUND

Increasing the appropriateness of use of upper gastrointestinal (GI) endoscopy is important to improve quality of care while at the same time containing costs. This study explored whether detailed explicit appropriateness criteria significantly improve the diagnostic yield of upper GI endoscopy.

METHODS

Consecutive patients referred for upper GI endoscopy at 6 centers (1 university hospital, 2 district hospitals, 3 gastroenterology practices) were prospectively included over a 6-month period. After controlling for disease presentation and patient characteristics, the relationship between the appropriateness of upper GI endoscopy, as assessed by explicit Swiss criteria developed by the RAND/UCLA panel method, and the presence of relevant endoscopic lesions was analyzed.

RESULTS

A total of 2088 patients (60% outpatients, 57% men) were included. Analysis was restricted to the 1681 patients referred for diagnostic upper GI endoscopy. Forty-six percent of upper GI endoscopies were judged to be appropriate, 15% uncertain, and 39% inappropriate by the explicit criteria. No cancer was found in upper GI endoscopies judged to be inappropriate. Upper GI endoscopies judged appropriate or uncertain yielded significantly more relevant lesions (60%) than did those judged to be inappropriate (37%; odds ratio 2.6: 95% CI [2.2, 3.2]). In multivariate analyses, the diagnostic yield of upper GI endoscopy was significantly influenced by appropriateness, patient gender and age, treatment setting, and symptoms.

CONCLUSIONS

Upper GI endoscopies performed for appropriate indications resulted in detecting significantly more clinically relevant lesions than did those performed for inappropriate indications. In addition, no upper GI endoscopy that resulted in a diagnosis of cancer was judged to be inappropriate. The use of such criteria improves patient selection for upper GI endoscopy and can thus contribute to efforts aimed at enhancing the quality and efficiency of care. (Gastrointest Endosc 2000;52:333-41).

Appropriateness of surgery for sciatica: reliability of guidelines from expert panels

STUDY DESIGN

Reliability study of guidelines development.

OBJECTIVE

To compare criteria for low back surgery between two expert panels.

BACKGROUND

Reliability of expert panels for determining appropriateness of indications for surgical procedures has heretofore received little attention.

METHODS

Two multidisciplinary expert panels of similar composition were convened, in the United States and in Switzerland, to evaluate the appropriateness of 720 distinct clinical scenarios involving sciatica. Each indication was assigned to a category of appropriate, uncertain, and inappropriate. The appropriateness of the 720 theoretical scenarios were compared between the two panels, and both sets of criteria were applied to two series of actual cases.

RESULTS

Seventy-nine percent (n = 566) of the 720 theoretical indications were assigned to identical categories of appropriateness by both panels (kappa = 0.63; P < 0.001). Only 2 of the 720 scenarios elicited frank disagreement. The percentage of the 720 indications that were considered appropriate differed between the two panels (U.S.: 3%; Swiss: 11%, P < 0.001), as did the percentage of intrapanel agreement for indications (U.S.: 51%, Swiss: 64%, P < 0.001). When the same theoretical scenarios were matched with two series of actual cases (n = 181 and 149) agreement was moderate (kappa = 0.46) to fair (kappa = 0.30).

CONCLUSION

There was substantial agreement on the appropriateness of surgery for theoretical cases of sciatica between independent expert panels from two countries. A better understanding of discordant ratings, especially for actual cases, should precede attempts at transposing recommendations emanating from a panel in one country to another.

Explaining drug spending trends: does perception match reality?

Several recent studies have made clear that drug expenditures are rising more rapidly than other health care spending. What has not been clear, however, is how much drug spending is driven by price rather than volume and whether volume increases are appropriate. This DataWatch takes a closer look at the components and drivers of drug spending using large claims databases from managed care and employer-sponsored health benefit plans. In both environments this study found volume, not price, to be the largest driver of drug spending for seven diseases studied. For four of the diseases, we review the clinical issues that may have influenced volume growth.

A clinically detailed risk information system for cost

The authors discuss a system that describes the resources needed to treat different subgroups of the population under age 65, based on burden of disease. It is based on 173 conditions, each with up to 3 severity levels, and contains models that combine prospective diagnoses with retrospectively determined elements. We used data from four different payers and standardized the cost of most services. Analyses showed that the models are replicable, are reasonably accurate, explain costs across payers, and reduce rewards for biased selection. A prospective model with additional payments for birth episodes and for serious problems in newborns would be an effective risk adjuster for Medicaid programs.

Appropriateness of gastrointestinal endoscopy: risk of complications

The balance between risks and benefits of gastrointestinal endoscopy for a given patient is essential in defining the appropriate use of endoscopic procedures. The current literature suggests that gastrointestinal endoscopy infrequently results in major procedure-related morbidity and mortality, while cardio-respiratory events occur commonly. However, true complication rates may be underestimated due to inconsistencies in the types of complications reported. No formal reporting requirements exist, and most of the published studies on complications come from centres with highly-skilled endoscopists.

Delphi-panel analysis of appropriateness of high-dose therapy and bone marrow transplants in chronic myelogenous leukemia in chronic phase

BACKGROUND

It is uncertain which people with chronic myelogenous leukemia (CML) in chronic phase should receive conventional treatment (interferon and/or chemotherapy) versus high-dose therapy and a bone marrow transplant. There are no randomized trials comparing these approaches and analyses of data from non-randomized studies are complex, contradictory without sufficient detail to allow subject-level treatment decisions.

OBJECTIVE

Determine appropriateness of high-dose therapy and bone marrow transplants in persons with CML in chronic phase with specific features. Develop a treatment algorithm. PANELISTS: nine leukemia experts from diverse geographic sites and practice settings.

EVIDENCE

Boolean MEDLINE searches of chronic myelogenous leukemia and chemotherapy and/or transplants.

CONSENSUS PROCESS

We used a modified Delphi-panel group judgment process. Age, prognostic score, disease duration, and type of conventional therapy and response were permuted to define 90 clinical settings. Each panelist rated appropriateness of high-dose therapy and a transplant versus conventional therapy on a 9-point ordinal scale (1, most inappropriate, 9, most appropriate) considering three types of donors: (1) HLA-identical siblings; (2) alternative donors (HLA-matched related or unrelated people other than an HLA-identical sibling); and (3) autotransplants. An appropriateness index was developed based on median rating and amount of disagreement. Relationship of appropriateness indices to permuted clinical variables was considered by analysis of variance and recursive partitioning. Preference between donor types was analyzed by comparing mean appropriateness indices of similar settings and a treatment algorithm developed.

CONCLUSIONS

In people with CML in chronic phase and an HLA-identical sibling donor and in those with an alternative donor (but no HLA-identical sibling), a transplant was rated appropriate in those with a < or = partial cytogenetic response to interferon and uncertain or inappropriate in all other settings. Autotransplants were rated uncertain or inappropriate in all settings. Most of the variance in appropriateness ratings between different clinical settings was accounted for by response to interferon: complete versus < or = partial response. An HLA-identical sibling donor, when available, was always preferred to an alternative donor or autotransplant. In people without an HLA-identical sibling, an alternative donor was favored over an autotransplant at higher appropriateness indices and the converse at lower appropriateness indices.

Delphi-panel analysis of appropriateness of high-dose therapy and bone marrow transplants in adults with acute myelogenous leukemia in 1st remission

BACKGROUND

Despite considerable data, there is still controversy over which adults with acute myelogenous leukemia (AML) in 1st remission should receive high-dose therapy and a bone marrow transplant rather than conventional-dose chemotherapy. Analyses of data from randomized trials are complex, conclusions sometimes contradictory and results not sufficiently detailed to allow subject-level decisions.

OBJECTIVE

To determine appropriate use of high-dose therapy and bone marrow transplants in persons with AML in 1st remission with specific features. Develop a treatment algorithm. PANELISTS: Nine leukemia experts from diverse geographic sites and practice settings.

EVIDENCE

Boolean MEDLINE searches of acute myelogenous leukemia and chemotherapy and/or transplants.

CONSENSUS PROCESS

We used a modified Delphi-panel group judgment process. Age, WBC, cytogenetics and FAB-type were permuted to define 72 clinical settings. Each panelist rated appropriateness of high-dose therapy and a transplant versus conventional-dose chemotherapy on a nine-point ordinal scale (1, most inappropriate, 9, most appropriate) considering 3 types of donors: (1) HLA-identical siblings; (2) alternative donors (HLA-matched related or unrelated people other than an HLA-identical sibling); and (3) autotransplants. An appropriateness index was developed based on median rating and amount of disagreement. The relationship of appropriateness indices to the permuted clinical variables was considered by analysis of variance and recursive partitioning. Preference between donor types was analyzed by comparing mean appropriateness indices of comparable settings and a treatment algorithm developed.

CONCLUSIONS

In people with an HLA-identical sibling, this type of transplant was rated appropriate in those with unfavorable cytogenetics and uncertain in all other settings. In people without an HLA-identical sibling, an alternative donor transplant was rated appropriate in those < 30 years with unfavorable cytogenetics, uncertain in those > 30 years and unfavorable cytogenetics and inappropriate in all other settings. Autotransplants were rated appropriate in people with unfavorable cytogenetics and uncertain in all other settings. An HLA-identical sibling donor, when available, was always preferred to an alternative donor transplant or autotransplant. In people without an HLA-identical sibling, an autotransplant was almost always favored over an alternative donor transplant with the magnitude of preference inversely correlated with transplant appropriateness.

Delphi-panel analysis of appropriateness of high-dose therapy and bone marrow autotransplants in newly diagnosed multiple myeloma

There is controversy whether high-dose therapy and a bone marrow autotransplant or conventional chemotherapy is a better treatment for newly diagnosed multiple myeloma. Data from 1 comparative study and 1 randomized trial provide insufficient subject-level data to advise specific people whether to have an autotransplant. We analyzed appropriate use of high-dose therapy and bone marrow autotransplants in people with newly diagnosed, multiple myeloma using a modified Delphi-panel group judgment process. The panel consisted of 9 myeloma experts from diverse geographic sites and practice settings who reviewed Boolean MEDLINE searches of multiple myeloma and chemotherapy or autotransplants. The panel rated a metric of 64 clinical setting developed by permuting age, performance score, disease-stage and disease-related prognostic variables and response to initial therapy. Each panelist rated appropriateness of high-dose therapy and an autotransplant versus conventional-dose chemotherapy on a 9-point ordinal scale (1, most inappropriate, 9, most appropriate). An appropriateness index was developed based on median rating and amount of disagreement. Relationship of appropriateness indices to the permuted clinical variables was considered by analysis of variance and recursive partitioning. Autotransplants were rated appropriate in persons <55 years old with stage 3 disease and a complete or partial response or stable disease after initial chemotherapy, inappropriate in persons with stage 1 or 2 disease, a performance score <70% and a complete or partial response or stable disease after initial chemotherapy and uncertain in all other settings.

The assessment of appropriate indications for laminectomy

We have developed criteria to determine the appropriate indications for lumbar laminectomy, using the standard procedure developed at the RAND corporation and the University of California at Los Angeles (RAND-UCLA). A panel of five surgeons and four physicians individually assessed 1000 hypothetical cases of sciatica, back pain only, symptoms of spinal stenosis, spondylolisthesis, miscellaneous indications or the need for repeat laminectomy. For the first round each member of the panel used a scale ranging from 1 (extremely inappropriate) to 9 (extremely appropriate). After discussion and condensation of the results into three categories laminectomy was considered appropriate in 11% of the 1000 theoretical scenarios, equivocal in 26% and inappropriate in 63%. There was some variation between the six categories of malalignment, but full agreement in 64% of the hypothetical cases. We applied these criteria retrospectively to the records of 196 patients who had had surgical treatment for herniated discs in one Swiss University hospital. We found that 48% of the operations were for appropriate indications, 29% for equivocal reasons and that 23% were inappropriate. The RAND-UCLA method is a feasible, useful and coherent approach to the study of the indications for laminectomy and related procedures, providing a number of important insights. Our conclusions now require validation by carefully designed prospective clinical trials, such as those which are used for new medical techniques.

Delphi-panel analysis of appropriateness of high-dose therapy and bone marrow transplants in adults with acute lymphoblastic leukemia in first remission

BACKGROUND

There is controversy over whether high-dose therapy and a bone marrow transplant is better than conventional-dose chemotherapy in adults with acute lymphoblastic leukemia (ALL) in first remission. This decision may depend on which type of donor is available: an HLA-identical sibling, an alternative donor transplant (HLA-matched related or unrelated people other than HLA-identical siblings), or autotransplant.

OBJECTIVE

To determine the appropriate use of high-dose therapy and bone marrow transplants in ALL in first remission. Develop a treatment algorithm. PANELISTS: Nine leukemia experts from diverse geographic sites and practice settings.

EVIDENCE

Boolean MEDLINE searches of acute lymphoblastic leukemia and chemotherapy and/or transplants.

CONSENSUS PROCESS

We used a modified Delphi-panel group judgment process. Age, white blood cell (WBC) count, cytogenetics and immune type were permuted to define 48 clinical settings. Each panelist rated appropriateness of high-dose therapy and a transplant versus conventional-dose chemotherapy on a 9-point ordinal scale (1, most inappropriate; 9, most appropriate) considering three types of donors: (1) HLA-identical siblings; (2) alternative donors; and (3) autotransplants. An appropriateness index was developed based on median rating and amount of disagreement. Relationship of appropriateness indices to the permuted clinical variables was considered by analysis of variance and recursive partitioning. Preference between donor types was analyzed by comparing mean appropriateness indices of comparable settings and a treatment algorithm was developed.

CONCLUSIONS

In people with an HLA-identical sibling donor, transplants were rated appropriate in those with unfavorable cytogenetics and uncertain in all other settings. An HLA-identical sibling donor was always preferred to an alternative donor or autotransplant. In people without an HLA-identical sibling but with an alternative donor, this type of transplant was rated appropriate in those with unfavorable cytogenetics. However, an autotransplant was preferred over an alternative donor transplant in all other settings where a transplant was rated uncertain. In people without an HLA-identical sibling or alternative donor, autotransplants were rated uncertain in all settings except in those with not unfavorable cytogenetics, WBC < 100 x 10(9) l(-1) and T- or pre-B-cell type where they were rated inappropriate.

The development of indications for the preoperative use of recombinant erythropoietin

OBJECTIVE

To develop indications for the preoperative use of recombinant erythropoietin (rHuEPO) alone and in conjunction with preoperative autologous donation (PAD).

DESIGN

A 2-round modified Delphi-consensus process.

PARTICIPANTS

Nine physicians representing multiple clinical specialties, practice environments and geographic locations.

METHOD

From evidence tables and a literature summary (MEDLINE database from January 1985 to August 1996) provided and using the RAND-UCLA appropriateness method, the physicians developed 264 indications for the preoperative use of rHuEPO by permuting 7 clinical factors (age, history of transfusion or antibody incompatibility, hemoglobin level, anemia of chronic disease, expected blood loss, presence of cardiovascular or cardiopulmonary disease and patient anxiety). These indications were rated on a 9-point appropriateness scale. Median scores and measures of agreement were determined.

OUTCOME MEASURES

The significance of cost constraints or cost and blood supply constraints and the impact of each clinical factor on the ratings as judged by statistical analysis.

RESULTS

Of the 264 indications, 54% were rated appropriate, 18% uncertain and 28% inappropriate. Expected blood loss had the greatest impact on the ratings (high expected blood loss had a 5.9 point more appropriate rating on the 9-point scale than low expected blood loss [p < 0.0001]). Preoperative hemoglobin level also significantly influenced the ratings (p < 0.0001). Compared with the clinical context, the ratings under the cost constraint were 1.0 less appropriate (p < 0.0001) for rHuEPO alone and 1.2 less appropriate for rHuEPO and PAD (p < 0.0001). The ratings for patients with moderate expected blood loss were significantly influenced by the cost constraint (less appropriate).

CONCLUSIONS

Expected blood loss and preoperative hemoglobin level were the best indicators of rHuEPO appropriateness. Different contexts modify the appropriateness ratings of an expensive drug like rHuEPO.

Efficacy of lumbar discectomy and percutaneous treatments for lumbar disc herniation

The changing health care environment necessitates careful re-evaluation of all costly elective procedures. Low back surgery is a typical example. This article reviews the current literature addressing the efficacy of surgery and invasive percutaneous treatments for discogenic sciatica. It also discusses the prospects for the continuation of reimbursement for these procedures under a system of managed health care. Relevant articles were identified using the MEDLINE and Current Contents databases, from bibliographies of articles identified from these databases, from recommendations of experts in the field, and from the Canadian Cochrane++ Collaboration. The review includes randomized clinical trials, meta-analyses, published practice guidelines and large case series. The literature is classified and discussed in these quality strata. The review includes 9 randomized trials, 6 meta-analyses or review articles, one evidence-based practice guideline, 38 surgical case series and 35 additional references. Though incomplete, the existing evidence indicates that open discectomy shortens the duration of discogenic sciatica in selected patients. Neurologic outcomes are similar in operated and unoperated patients. Predominant leg pain, evidence of nerve root tension and concordant symptoms and imaging findings, are associated with favorable surgical results. Chemonucleolysis is also associated with more rapid pain relief than conservative treatment, but provides less certain benefit than standard discectomy. Available data on other percutaneous disc treatments do not currently support a statement on efficacy. Various percutaneous techniques are available but there is no solid scientific evidence of efficacy. The benefits of open discectomy, principally reduced duration of pain, appear to justify its use in carefully selected patients when discogenic sciatica fails to improve with conservative measures. Though elective, the procedure will probably continue to be available under managed care, but with increasing scrutiny of operative indications.

Reliability of panel-based guidelines for colonoscopy: an international comparison

BACKGROUND

This study examined the reliability of explicit guidelines developed using the RAND-UCLA appropriateness method.

METHODS

The appropriateness of over 400 indications for colonoscopy was rated by two multispecialty expert panels (United States and Switzerland). A nine-point scale was used, which was consolidated into three categories of appropriateness: appropriate, uncertain, inappropriate. The distribution of appropriateness ratings between the two panels and the intrapanel and interpanel agreement for categories of appropriateness were calculated for all possible indications. Similar statistics were calculated for a series of 577 primary care patients referred for colonoscopy in Switzerland.

RESULTS

Over 80% of all indications (348) could be directly compared. The proportions of indications classified as appropriate, uncertain, or inappropriate were 28.4%, 24.7%, 46.6% and 33.0%, 23.0%, 44.0% for the U.S. and the Swiss panels, respectively. Interpanel agreement was excellent for all the possible indications (kappa value: 0.75) and lower for actual cases (kappa value: 0.51) because of lower agreement for the most frequently encountered indications.

CONCLUSIONS

Good agreement between the two sets of criteria was found, pointing to the reliability of the method. Partial disagreement occurred essentially for a few, albeit frequently encountered, indications for use of colonoscopy in cases of uncomplicated lower abdominal pain or constipation.

Gastrointestinal illness in managed care: healthcare utilization and costs

Identification of inefficiencies is a first step to improving the quality of gastrointestinal (GI) care at the most reasonable cost. This analysis used administrative data to examine the healthcare utilization and associated costs of the management of GI illnesses in a 2.5 million-member private managed care plan containing many benefit designs. An overall incidence of 10% was found for GI conditions, with a preponderance in adults (patients older than 40 years) and women. The most frequently occurring conditions were abdominal pain, nonulcer peptic diseases, lower GI tract diseases, and other GI tract problems. These conditions, along with gallbladder/biliary tract disease, were also the most costly. Claims submitted for care during GI episodes averaged $17 per member per month. Increasing severity of condition was associated with substantial increases in utilization and costs (except for medication use). For most GI conditions, approximately 40% of charges were for professional services (procedures, tests, and visits) and 40% of charges were for facility admissions. The prescription utilization analysis indicated areas where utilization patterns may not match accepted guidelines, such as the low use of anti-Helicobacter pylori therapy, the possible concomitant use of nonsteroidal anti-inflammatory drugs in patients with upper GI diseases, and the use of narcotics in treating patients with lower GI disease and abdominal pain. Also, there was no clear relationship between medication utilization and disease severity. Thus, this analysis indicated that GI disease is a significant economic burden to managed care, and identified usage patterns that potentially could be modified to improve quality of care.

Appropriateness of upper gastrointestinal endoscopy: comparison of American and Swiss criteria

OBJECTIVE

Examine the reproducibility of the RAND method for developing criteria for the appropriateness of medical procedures.

DESIGN

Comparison of two sets of explicit criteria for appropriateness of upper gastrointestinal (UGI) endoscopy, developed by separate expert panels from two countries.

SETTING

United States, Switzerland.

STUDY PARTICIPANTS

National experts from different medical specialties involved in the referral or application of UGI endoscopy.

INTERVENTIONS

Each panel was presented with about 500 clinical scenarios (indications) that were rated on a nine-point scale as to the appropriateness of performing UGI endoscopy for a patient with that clinical presentation.

MAIN OUTCOME MEASURES

(1) distribution of appropriateness ratings and intrapanel agreement categories between the two panels, (2) between-panel agreement of assigning appropriateness for comparable indications and, (3) percentage of indications with major between-panel differences.

RESULTS

Ratings for 2/3 of indications could be compared. The Swiss panel showed higher intrapanel agreement (54.6% versus 46.2%, P = 0.002). Seventy-eight per cent of comparable indications were assigned to identical categories of appropriateness by both panels (kappa = 0.76, P < 0.001). For 93% of the 376 comparable indications, there were no major interpanel differences.

CONCLUSION

Separate expert panels in different countries, using a standardized methodology, produce criteria for appropriateness of medical procedures that are similar. Given the resources being invested throughout the world in developing criteria and guidelines, international collaboration in seeking optimal use of limited health care resources should be intensified.

The medical appropriateness of tympanostomy tubes proposed for children younger than 16 years in the United States

OBJECTIVE

To describe the clinical reasons tympanostomy tubes are proposed for children and to assess their appropriateness.

DESIGN

Analysis of data previously collected prospectively by a national utilization review (UR) firm during a two-step UR process to assess the medical appropriateness of tympanostomy tube placement. Nurses interviewed otolaryngologists' and primary care physicians' office staff to collect clinical data. For a randomly selected subsample of cases found inappropriate, we reviewed subsequent interviews of the otolaryngologists by physician reviewers, who looked for possible extenuating clinical circumstances or additional clinical data that might have changed the appropriateness category.

SETTING

Otolaryngologists' practices from 49 states and the District of Columbia.

PATIENTS

All 6611 children younger than 16 years who were insured by three clients of the UR firm and whose proposal to receive tympanostomy tubes were reviewed by this system from January 1, 1990, through July 31, 1991. The insurance companies in the study insured 5.6 million Americans at the time of the study.

MAIN OUTCOME MEASURE

The medical appropriateness of tympanostomy tube surgery according to explicit criteria developed by an expert panel using the RAND/University of California-Los Angeles modified Delphi method.

RESULTS

A total of 6429 (97%) of the cases were proposed for recurrent acute otitis media, otitis media with effusion, or both. Making generous clinical assumptions, 41% of the proposals for these reasons had appropriate indications, 32% had equivocal indications, and 27% had inappropriate ones. Considering the additional information available from the subsample review, the proportion appropriate was 42%, equivocal 35%, and inappropriate 23%.

CONCLUSION

About one quarter of tympanostomy tube insertions for children in this study were proposed for inappropriate indications and another third for equivocal ones.

Methods to assess the quality of care for elderly patients with pneumonia

During the past several years, interest has focused on the cost and quality of health care. The elderly represent an important population to study since they are less resilient to the effects of acute illness such as pneumonia, and if they receive improper care, these patients will be more sensitive to that poor care and more likely demonstrate measurable ill effects. This article examines several methods to assess quality. The traditional method addresses the process of care or whether the patient received the appropriate care for his or her particular condition. Process of care evaluations either rely upon the subjective judgment of expert clinicians or use sets of explicitly defined criteria. A second quality assessment method examines the patient outcome, ie, regardless of the care received, did the patient get well? Each of these methods has strengths and limitations. Based upon this understanding, the article will describe a multimodality approach to the assessment of quality.

Preventable deaths: who, how often, and why?

If the quality of care provided by a hospital affects its death rate, then some deaths must be preventable. We have developed a new method to investigate this issue and have reviewed 182 deaths from 12 hospitals (6 high outliers and 6 low outliers for death rate) for three conditions (cerebrovascular accident, pneumonia, or myocardial infarction). The investigators prepared a dictated summary of each patient's hospital course. Then, at least three physicians reviewed each summary and independently judged whether the death could have been prevented. Using a majority rules criterion (at least two of three physicians agreed), we found that 27% of the deaths might have been prevented. Using a unanimity criterion (all three physicians independently agreed), we found a 14% rate of probably preventable deaths. Patients whose deaths were probably preventable were younger (74.7 compared with 78.6 years, P less than 0.05), less often demented (12% compared with 26%, P less than 0.05), and less severely ill (mean Acute Physiology and Chronic Health Evaluation score, 15.6 compared with 21.2; P less than 0.001) than patients whose deaths were nonpreventable. The physicians also listed causes for each probably preventable death; nine reasons encompassed almost all of them. For myocardial infarction, preventable deaths reflected errors in management. For cerebrovascular accident, however, deaths primarily reflected errors in diagnosis. The severity of illness can help a hospital retrospectively identify probably preventable deaths. In the group of patients who died, 42% of those with a low severity of illness had probably preventable deaths as compared to 11% admitted with a high severity of illness. We found that a significant number of hospital deaths might have been prevented. Our findings were based on a new method that needs further testing to substantiate its validity. These findings also need replication before they can be generalized to other hospitals.

Hospital inpatient mortality. Is it a predictor of quality?

Various potential measures of quality of care are being used to differentiate hospitals. Last year, on the basis of diagnostic and demographic data, the Health Care Financing Administration identified hospitals in which the actual death rate differed from the predicted rate. We have developed a similar model. To understand why there are high-outlier hospitals (in which the actual death rate is above the predicted one) and low-outlier hospitals (in which the actual death rate is below the predicted one), we reviewed 378 medical records from 12 outlier hospitals treating patients with one of three conditions: cerebrovascular accident, myocardial infarction, and pneumonia. After adjustment for the severity of illness, the death rate in the high outliers exceeded that predicted from the severity of illness alone by 3 to 10 percent, and in the low outliers, the actual death rate fell short of the severity-adjusted predictions by 10 to 15 percent (P less than 0.01). Reviews of the process of care using 125 criteria revealed no differences between the high and low outliers. However, detailed reviews by physicians of the records of patients who died during hospitalization revealed a higher rate of preventable deaths in the high outliers than in the low outliers. For the three conditions studied, we project that 5.7 percent of a standard cohort of patients admitted to the high-outlier hospitals would have preventable deaths, as compared with 3.2 percent of patients admitted to the low-outlier hospitals (P less than 0.05). A meaningful comparison of hospital death rates requires adjustment for severity of illness. Our findings indicate that high-outlier hospitals care for sicker patients. However, these same hospitals or their medical staffs may also provide poorer care. Our results need confirmation before death-rate models can be used to screen hospitals.

Adjusted hospital death rates: a potential screen for quality of medical care

Increased economic pressure on hospitals has accelerated the need to develop a screening tool for identifying hospitals that potentially provide poor quality care. Based upon data from 93 hospitals and 205,000 admissions, we used a multiple regression model to adjust the hospitals crude death rate. The adjustment process used age, origin of patient from the emergency department or nursing home, and a hospital case mix index based on DRGs (diagnostic related groups). Before adjustment, hospital death rates ranged from 0.3 to 5.8 per 100 admissions. After adjustment, hospital death ratios ranged from 0.36 to 1.36 per 100 (actual death rate divided by predicted death rate). Eleven hospitals (12 per cent) were identified where the actual death rate exceeded the predicted death rate by more than two standard deviations. In nine hospitals (10 per cent), the predicted death rate exceeded the actual death rate by a similar statistical margin. The 11 hospitals with higher than predicted death rates may provide inadequate quality of care or have uniquely ill patient populations. The adjusted death rate model needs to be validated and generalized before it can be used routinely to screen hospitals. However, the remaining large differences in observed versus predicted death rates lead us to believe that important differences in hospital performance may exist.