Laboratory Testing, Calcium: Task A -- Draft Report
The recent trend of linking medical providers with patients to create integrated health care networks has helped reduce clinical practice variation. Consensus guidelines and disease management strategies have standardized the medical approach to many common disorders. This reduction in clinical practice variation has accentuated the effects of laboratory variation. Unfortunately, the developers of medical guidelines have assumed that all laboratories function well and all test results are comparable. Medical guidelines seldom contain any information about the performance characteristics for key tests used in the diagnostic decision process. Guidelines typically have specific thresholds or acceptable ranges, such as 8.9 to 10.1 mg/dL of calcium in tests to diagnose hypercalcemia, without any reference to measurement methodology or measurement standardization. Consequently, there is a false sense of security that the health care system assures adequate quality for laboratory tests. Analytic bias is a key parameter affecting the number of patients passing decision thresholds in practice guidelines. The Food and Drug Administration (FDA) requires that new tests perform equivalent to previously approved methods but does not require metrological traceability to reference methods. The Centers for Medicare & Medicaid Services (CMS) requires that laboratories be certified according to the Clinical Laboratory Improvement Act (CLIA 88) regulations. However, the CLIA performance limits for proficiency tests are very wide, and most tests are evaluated relative to peer groups. This practice allows large between-lot differences within methods and large between-method variations. This study investigates the potential impact on health care costs from analytic bias in tests to measure serum calcium levels. Hypercalcemia is a medical condition caused by various disorders—most commonly hyperparathyroidism and cancer. The signs and symptoms of hypercalcemia are nonspecific; therefore, the clinician is very dependent on accurate laboratory measurements for detecting and evaluating this disorder. Medical guidelines recommend that hypercalcemia be confirmed with follow-up procedures, such as intact parathyroid hormone (PTH) measurement, chest X-rays, 24-hour urinary calcium measurement, ionized calcium measurement, and thyroid imaging. Based on analysis of over 89,000 patients receiving serum calcium tests at the Mayo Clinic in 1998–1999, we find that the number of follow-up procedures, and hence health care costs, is directly related to initial calcium test values. It is estimated that the cost impact associated with an analytical bias of 0.1 mg/dL could range from $8 to $31 per patient (receiving a calcium test). For an analytical bias of 0.5 mg/dL, which was the approximate upper bound identified during interviews with laboratory managers and equipment manufacturers, the potential health care cost increase ranged from $34 to $89 per patient having a calcium test. With approximately 24 million patients per year receiving initial calcium screenings, the potential economic impacts range from $400 million to $1.3 billion per year for analytic biases of 0.1 and 0.5 mg/dL, respectively.