Evaluation of Selected Laboratory Components of a Comprehensive Periodic Health Evaluation for Veterans With Spinal Cord Injury and Disorders

Article Outline

• Abstract
• Methods
  • Study Population
  • Data Collection
  • Outcome Measures
  • Statistical Analysis
• Results
  • Patient Characteristics
  • Laboratory Test Associations With Patient Demographics and Neurologic Status
  • Diagnostic and Therapeutic Yields
• Discussion
  • Study Limitations
• Conclusions
• Acknowledgment
• References
• Copyright

Abstract 

Choi H, Binder DS, Oropilla ML, Bernotus EE, Konya D, Nee MA, Tammaro EA, Sabharwal S. Evaluation of selected laboratory components of a comprehensive periodic health evaluation for veterans with spinal cord injury and disorders.

Objective

To evaluate selected laboratory components of a comprehensive periodic health evaluation program for patients with spinal cord injury and disorders (SCI/D).

Design

A retrospective study.

Setting

A Department of Veterans Affairs spinal cord injury center.

Participants

Community-dwelling male veterans with SCI/D (N=350).

Interventions

Not applicable.

Main Outcome Measures

Proportion of laboratory tests that resulted in new diagnoses (diagnostic yield) and proportion of laboratory tests that resulted in changes in management (therapeutic yield).

Results

Although abnormality rates for many routine laboratory tests were high (up to 31.5%), diagnostic and therapeutic yields were low (<1.5%), with the exception of glucose (therapeutic yield, 3.4%) and lipid tests (up to a 4.1% diagnostic and 15.2% therapeutic yield).

Conclusions

Our data revealed that diagnostic and therapeutic yields for many laboratory components of the annual PHE program for veterans with SCI/D were low, consistent with findings in the general ambulatory population. Further data collection, particularly prospective longitudinal data, may help optimize the selection and frequency of laboratory tests performed as part of this program.

Key Words:  Cross-sectional studies , Preventive medicine , Rehabilitation , Spinal cord injuries , Veterans

GUIDELINES FROM THE U.S. PREVENTIVE Services Task Force (USPSTF) do not generally advocate routine laboratory and other investigations for asymptomatic healthy adults during periodic health evaluations (PHEs).¹, ², ³ Similar positions are endorsed by the American College of Physicians,⁴ the American Medical Association,⁵ and the Canadian Task Force on Periodic Health Examination.⁶, ⁷ With the exception of lipid and glucose tests, laboratory screening infrequently identifies treatable disorders.⁸, ⁹, ¹⁰, ¹¹, ¹² Prevalence of disease in asymptomatic patients is known to be low, increasing the risk of false-positive results.⁸ Abnormal results are also frequently ignored or are followed by additional investigations or consultations without implementation of treatment.⁸, ¹⁰ In its last recommendations published in 1996, the USPSTF only endorsed the Papanicolaou smear, periodic (not necessarily annual) nonfasting total cholesterol levels, and screening in certain age groups or risk profiles for colorectal cancer, breast cancer, and chlamydia.³ Routine blood tests such as the complete blood count (CBC) or basic chemistry profile in asymptomatic adults were not advocated. The literature on many examinations, including CBCs, chemistry profiles, and other individual or panel laboratory tests in general ambulatory care,¹³, ¹⁴ preoperative adult patients,⁹ or severely impaired residents of nursing homes,¹⁵ is likewise unsupportive of routine periodic testing in asymptomatic adults.

It is not known, however, whether these recommendations are generalizable for persons with spinal cord injury and disorders (SCI/D), who are at higher risk to develop certain complications such as those of the respiratory and urologic systems,¹⁶ diabetes mellitus,¹⁷ and hepatitis C viral infection¹⁸ more frequently than the non-SCI/D population. Of particular concern in the population with SCI/D is the potential for absent signs or symptoms because of sensory deficits, with the possibility of delayed disease detection. Data describing the utility of routine laboratory tests (ie, the rate of establishing clinically significant new diagnoses or implementing treatments because of the tests) in the SCI/D population, such as those performed as part of annual multiphasic (ie, “multiple test”)¹⁹ PHEs, however, have not been published to our knowledge.

The main objective of the present study was to determine retrospectively the rates at which new diagnoses had been made and therapeutic interventions implemented following the results of selected routine laboratory tests performed as part of an annual comprehensive PHE program for veterans with SCI/D at the U.S. Department of Veterans Affairs (VA) Boston Healthcare System (BHCS) Spinal Cord Injury Center during 2001 to 2003.²⁰ We were also interested in examining the relations between laboratory results and demographic factors such as age, duration post-SCI/D, and injury severity to explore if any factors might be independently associated with laboratory results, under the supposition that some patterns of associations between test result variability and demographic variables might infer that guidelines for the SCI/D community may significantly depart from those for the general population.

Back to Article Outline

Methods 

Study Population 

The study population consisted of all community-dwelling male veteran patients with chronic SCI/D who were 18 years of age older, registered with the VA BHCS SCI/D service, and who had completed at least 1 PHE during calendar years 2001 to 2003. For patients completing more than 1 PHE during the time period, only the most recent PHE was used. Women, who comprised less than 1% of our otherwise eligible study population, were excluded. Additional exclusion criteria included history of SCI/D of less than 1 year in duration; institutionalization; or a principal neurologic diagnosis of multiple sclerosis, transverse myelitis, or amyotrophic lateral sclerosis. The study was approved by the VA BHCS Institutional Review Board.

Data Collection 

Data were abstracted retrospectively from the VA BHCS electronic medical record, including patient demographics, duration post-SCI/D, American Spinal Injury Association (ASIA) Impairment Scale,²¹ and setting of presentation (ie, as an outpatient or short-term inpatient). All the laboratory tests and follow-up clinical decision making reviewed in this study had occurred antecedent to the conception of this study. Patients were categorized a priori into groups by motor and autonomic severity to examine whether motor function capabilities (eg, full motor function of the upper limbs) and/or intactness of connections between the higher central nervous system and the major sympathetic outflows of the autonomic nervous system²¹ might also affect laboratory values. Motor function was categorized into 4 subgroups: motor complete (ie, ASIA grade A or B) C1 to T1, motor incomplete (ie, ASIA grade C or D) C1 to T1, motor complete T2 to L5, and motor incomplete T2 to L5. Because some T1-level patients could have less than full strength (Medical Research Council grade <5)²² in their distal upper limbs, patients with a T1 neurologic level were placed into the category that included patients with tetraplegia. Subjects were also divided into 4 subgroups based on intactness of autonomic connections, that is, whether the injury was above or below the sympathetic splanchnic outflow: C1 to T6 complete (ie, ASIA grade A), C1 to T6 incomplete (ie, ASIA grades B–D), T7 to L5 complete, and T7 to L5 incomplete.

For each patient, we reviewed select laboratory data from the PHE (ie, CBC, chemistry panel, fasting glucose, liver function tests, fasting lipid profile) including results dating back 1 year (excluding the previous PHE) and for 6 months after the PHE, follow-up notes, consultation reports, and health care provider order records. New diagnoses established and treatments implemented by the primary SCI/D clinicians or consultants as a result of PHE laboratory tests or follow-up tests within 6 months of the PHE were included. If more than 1 test contributed to a new diagnosis or therapy, then each separate laboratory abnormality was credited with a new diagnosis or therapy. For instance, if a patient had both an elevated total cholesterol and elevated low-density lipoprotein cholesterol (LDL-C) level and was therefore started on a lipid-lowering agent, each test would be considered to have yielded a therapeutic intervention. Normal laboratory test results were defined by VA laboratory reference ranges contemporaneous to the test dates, which did not change from 2000 to 2004.

Outcome Measures 

The associations between demographic factors and laboratory values were explored by using multivariate analysis. The rates at which each laboratory test triggered new (previously unknown or undocumented), clinically significant diagnoses (ie, diagnostic yield) or changes in management (ie, therapeutic yield) were determined.¹¹ When therapeutic yields were noted to be high (ie, glucose and lipid tests), treatments were also subdivided post hoc into “education/counseling only” treatments and treatment “interventions” (eg, medication initiation, blood transfusion, or surgery). Other outcome measures included the proportion of laboratory tests with abnormal results and the rates at which the PHE resulted in follow-up test ordering or specialist consultation. Abnormal laboratory results were further defined as “mild” if laboratory values were within 10% of the upper or lower limits of the reference range or “severe” if outside of this range.¹¹ Treatments that were declined or not implemented were not counted toward the therapeutic yield. All the results of this study, including the diagnoses, therapeutic management decisions, and follow-up, were reviewed by a consensus panel of the lead authors (HC, DSB, MLO) and senior author (SS) of this study.

Statistical Analysis 

We used a multiway analysis of covariance (ANCOVA) model to detect independent associations of laboratory test results with the patient covariates age, duration post-SCI/D, SCI/D severity category, and inpatient or outpatient setting. For SCI/D severity category, only the category (ie, motor or autonomic) that accounted for the greatest amount of total variability in the model (ie, the greatest square of the correlation coefficient [R²] in the ANCOVA) was fitted into the model. When statistically significant in the multivariate model (2-tailed, P<.05), univariate relations between age and laboratory values for each SCI/D severity category were graphically explored post hoc by using locally fitted regression surface estimation plots smoothed by multivariate techniques (ie, locally weighted regression curves).²³, ²⁴ Multiple logistic regression analyses were conducted to test patient factors (eg, duration postinjury, age, SCI/D severity categories) as independent predictors of treatment. Therapeutic yield differences between supervising practitioners were compared by using the chi-square test. All results are expressed as means ± standard error of the mean (SEM), except where otherwise specified. The Stata statistical package ^a was used for all analyses.

Back to Article Outline

Results 

Patient Characteristics 

The clinical characteristics of the study population (N=350) are listed in table 1. The most prevalent neurologic level of injury was C6 (12%), followed by C4 (9.4%), C5 (9.1%), C7 (7.7%), T12 (6.6%), and C8 (5.1%). There were no patients with a sacral level of injury. Seventy-five (21.4%) PHEs were performed as inpatient admissions for multifactorial reasons such as transportation issues, social issues, or concomitant scheduling of other tests such as magnetic resonance imaging or colonoscopy, whereas the remainder of cases (n=275 [78.6%]) were performed as single-day outpatient visits.

Table 1. Study Population Demographics (N=350)

Demographics	Values
PHE date and setting of presentation, n (%)
2001	40(11.4)
2002	59(16.9)
2003	251(71.7)
Inpatient	75(21.4)
Outpatient	275(78.6)
Patient age(y)
Range	23–90
Mean ± SEM	57.1±0.8
Time interval since SCI/D(y)
Range	1–58
Mean±SEM	22.6±0.8
Ethnicity, n(%)
White	331(94.6)
African American	12(3.4)
Hispanic	4(1.1)
Other	3(0.9)
Etiology of injury, n(%)
Motor vehicle collision	133(38.0)
Fall	76(21.7)
Sports	37(10.6)
Medical	37(10.6)
Violence	34(9.7)
Trauma, other	33(9.4)
ASIA classification, n(%)
Grade A	170(48.6)
Grade B	32(9.1)
Grade C	34(9.7)
Grade D	114(32.6)
Motor category, n(%)
C1-T1; ASIA grade A, B	76(21.7)
C1-T1, ASIA grade C, D	90(25.7)
T2-L5, ASIA grade A, B	126(36.0)
T2-L5, ASIA grade C, D	58(16.6)
Autonomic category, n(%)
C1-T6, ASIA grade A	104(29.7)
C1-T6, ASIA grades B−D	117(33.4)
T7-L5, ASIA grade A	66(18.9)
T7-L5, ASIA grades B−D	63(18.0)

Laboratory Test Associations With Patient Demographics and Neurologic Status 

Not all patients had complete sets of laboratory results available. For unverifiable reasons, a small number of patients were missing some laboratory panels or portions thereof (ie, CBC, n=3; chemistry panel, n=1; liver function tests, n=6; lipid panel, n=11). The available laboratory data were evaluated for departures from Gaussian distributions by examining skewness, kurtosis, normal probability plots, and kernel density plots.²⁵ In some instances, outliers were removed post hoc and normality achieved or approached.

Associations between laboratory values and demographic characteristics were evaluated by using multiway ANCOVA. Time interval since onset of SCI/D was a significant factor in the univariate analyses for several laboratory test results (hematocrit, sodium, albumin; P<.05) but generally not a significant factor in the multivariate models. In our study, only the alanine aminotransferase (ALT) (P=.045) and aspartate aminotransferase (AST) (P=.073) data showed an independent association or trend toward association with duration postinjury. Statistically significant associations between laboratory data and age in the multivariate models, on the other hand, were noted for serum albumin, creatinine, hematocrit, ALT, blood urea nitrogen (BUN), AST, sodium, high-density lipoprotein cholesterol (HDL-C), and glucose (table 2, fig 1), consistent with previous reports for some of these laboratory tests in the SCI/D²⁶ and general ambulatory literature.²⁷, ²⁸, ²⁹, ³⁰ The overall adjusted squares of the correlation coefficients for the multivariate models, however, were generally very low (R²≤.245), meaning that little of the total variability (ie, ≤24.5%) could be explained by the covariates in our models. This begs the question whether additional demographic or health status factors (eg, use of medications with known effects on laboratory values) not yet incorporated into our models could increase the R² values and potentially weaken the observed independent associations between the laboratory values and demographic variables examined in our study. The greatest proportion of the explicable variability for most laboratory tests was because of age (data not shown).

Table 2. Results of Multiway ANCOVA

Abbreviations: AP, alkaline phosphatase; NA, not applicable; WBC, white blood cell count.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  Scatterplots of laboratory test values versus age with univariate regression lines and 95% confidence interval bands and locally weighted regression curves of laboratory test values versus age by SCI/D severity category; VA BHCS reference ranges in parentheses.

Diagnostic and Therapeutic Yields 

In the analysis of diagnostic and therapeutic yields, all laboratory test results were used (ie, outliers that had been removed from the ANCOVA were not excluded [table 3]). A total of 654 (11.2%) abnormal laboratory values resulting from 5865 laboratory tests were observed, and the proportions of abnormal values for individual laboratory tests were generally high. For example, we observed abnormal laboratory test result rates of 14% for sodium levels, 23.5% for glucose, 27.9% for total cholesterol, and 31.5% for hematocrit (see table 3), all grossly consistent with the SCI/D literature.³¹, ³², ³³, ³⁴ The lowest rate of abnormal laboratory test results was for creatinine at 1.1%; however, the underlying reasons for this are well described in the SCI/D literature.³⁵

Table 3. Diagnostic and Therapeutic Yields

Laboratory Test (n)	Abnormal Tests, n (%)	Overall New Diagnoses, n (%)	Overall New Therapies, n (%)	Case Severity ± Follow-Up	Abnormal Tests, n (% of cases)	New Diagnoses, n (% of cases)	New Therapies, n (% of cases)
WBC(346)	33(9.5)	0(0.0)	1(0.3)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	21(63.6) 0(0.0) 5(15.2) 7(21.2)	0(NA) 0(NA) 0(NA) 0(NA)	0(0.0) 0(0.0) 0(0.0) 1(100.0)
HCT(349)	110(31.5)	5(1.4)	5(1.4)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	68(61.8) 5(4.6) 24(21.8) 13(11.8)	0(0.0) 1(20.0) 0(0.0) 4(80.0)	1(20.0) 0(0.0) 0(0.0) 4(80.0)
Platelets(345)	14(4.1)	1(0.3)	1(0.3)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	2(14.3) 0(0.0) 5(35.7) 7(50.0)	0(0.0) 0(0.0) 0(0.0) 1(100)	0(0.0) 0(0.0) 0(0.0) 1(100)
Sodium(349)	49(14)	0(0.0)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	46(93.9) 3(6.1) 0(0.0) 0(0.0)	0(NA) 0(NA) 0(NA) 0(NA)	0(NA) 0(NA) 0(NA) 0(NA)
Potassium(349)	11(3.2)	0(0.0)	2(0.6)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	9(81.8) 0(0.0) 0(0.0) 2(18.2)	0(NA) 0(NA) 0(NA) 0(NA)	0(0.0) 0(0.0) 0(0.0) 2(100.0)
BUN(349)	21(6.0)	0(0.0)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	6(28.6) 0(0.0) 15(71.4) 0(0.0)	0(NA) 0(NA) 0(NA) 0(NA)	0(NA) 0(NA) 0(NA) 0(NA)
Creatinine(349)	4(1.1)	0(0.0)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	1(25.0) 0(0.0) 2(50.0) 1(25.0)	0(NA) 0(NA) 0(NA) 0(NA)	0(NA) 0(NA) 0(NA) 0(NA)
Glucose(349)	82(23.5)	1(0.3)	12(3.4)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	21(25.6) 4(4.9) 44(53.6) 13(15.9)	0(0.0) 0(0.0) 0(0.0) 1(100.0)	0(0.0) 1(8.3) 6(50.0) 5(41.7)
Albumin(344)	14(4.1)	0(0.0)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	9(64.3) 2(14.3) 2(14.3) 1(7.1)	0(0.0) 0(0.0) 0(0.0) 0(0.0)	0(0.0) 0(0.0) 0(0.0) 0(0.0)
ALT(344)	15(4.4)	1(0.3)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	6(40.0) 0(0.0) 4(26.7) 5(33.3)	0(0.0) 0(0.0) 0(0.0) 1(100.0)	0(NA) 0(NA) 0(NA) 0(NA)
AST(344)	8(2.3)	1(0.3)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	1(12.5) 1(12.5) 4(50.0) 2(25.0)	0(0.0) 1(100.0) 0(0.0) 0(0.0)	0(NA 0(NA) 0(NA) 0(NA)
AP(344)	28(8.1)	0(0.0)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	12(42.9) 0(0.0) 15(53.5) 1(3.6)	0(NA) 0(NA) 0(NA) 0(NA)	0(NA) 0(NA) 0(NA) 0(NA)
Total bilirubin(344)	14(4.1)	0(0.0)	0(0.0)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	3(21.4) 0(0.0) 11(78.6) 0(0.0)	0(NA) 0(NA) 0(NA) 0(NA)	0(NA) 0(NA) 0(NA) 0(NA)
Total cholesterol(338)	95(27.9)	14(4.1)	52(15.2)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	40(42.1) 11(11.6) 25(26.3) 19(20)	2(14.3) 2(14.3) 4(28.6) 6(42.8)	18(34.6) 9(17.3) 11(21.2) 14(26.9)
HDL-C(341)	65(9.1)	2(0.6)	5(1.5)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	35(53.9) 1(1.5) 24(36.9) 5(7.7)	0(0.0) 0(0.0) 0(0.0) 2(100.0)	2(40.0) 1(20.0) 0(0.0) 2(40.0)
LDL-C(336)	26(7.7)	6(1.8)	14(4.2)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	6(23.1) 5(19.2) 11(42.3) 4(15.4)	1(16.7) 1(16.7) 2(33.3) 2(33.3)	3(21.4) 3(21.4) 5(35.8) 3(21.4)
Triglycerides(341)	65(19.1)	9(2.6)	29(8.5)	Mild cases, − F/U Mild cases, + F/U Severe cases, − F/U Severe cases, + F/U	12(18.5) 7(10.8) 29(44.6) 17(26.1)	2(22.3) 1(11.1) 3(33.3) 3(33.3)	3(10.3) 3(10.3) 11(38.0) 12(41.4)

Abbreviations: F/U, follow-up; HCT, hematocrit; −, indicates that no follow-up tests or consults were ordered; +, indicates that follow up tests and/or consults were ordered.

Legend: mild cases, within 10% of upper or lower limits of the reference range; severe cases, ≥10% of the upper or lower limits of the reference range.

Although the proportions of abnormal laboratory values were generally high, the abnormalities were often mild (n=337 [51.5%]) and new clinically significant diagnoses were rarely made (total of 40 [0.7%] new diagnoses out of 5865 laboratory tests). The most common new diagnosis was dyslipidemia (n=14), followed by hypertriglyceridemia (n=9). Total cholesterol resulted in a 4.1% diagnostic yield, which was the highest rate for any of the tests among those studied, followed by 2.6% for triglycerides (see table 3). Thirty-two (80%) of the 40 new diagnoses were incited by glucose or lipid testing. Many of the laboratory tests (eg, sodium, potassium, albumin, creatinine) did not result in any significant new diagnoses. Of 136 abnormal test results that were followed up with further tests and/or consults, only 26 (19.1%) resulted in new diagnoses. The 110 abnormal tests that did not yield a new diagnosis included 77 severely abnormal results and involved follow-ups that included colonoscopies in 3 patients (with incidentally noted low hematocrits) who had already had their routine surveillance colonoscopies performed at the appropriate earlier dates.

Therapeutic yields were likewise low (see table 3). Overall, a total of 121 (2.1%) interventions resulting from 5865 laboratory tests were observed. Together, lipid and glucose tests accounted for 112 (92.6%) of the interventions, with 44 (36.4%) involving the initiation or adjustment of medications because of dyslipidemia and 4 (3.3%) because of suboptimal glycemic control. The remainder of the interventions because of abnormal glucose or lipid tests involved educational or counseling efforts without other interventions (52.9%). These corresponded to a 15.2% therapeutic yield for the lipid tests collectively and 3.4% for glucose. Seventy-five abnormal laboratory tests (including 53 severely abnormal results) were followed by additional tests and/or consults without resulting in a therapeutic intervention.

Excluding glucose and lipid tests, the other laboratory tests resulted in 9 total interventions. These included dietary counseling for mild abnormalities in potassium levels (n=2), an antibiotic incited by leukocytosis, an immune globulin injection for thrombocytopenia in a patient already being followed by the hematology/oncology service for immune thrombocytopenic purpura, the initiation of vitamins for anemia (n=2), a proton pump inhibitor after the discovery of gastritis on esophagoduodenoscopy incited by anemia, and 2 nonemergent transfusions for anemia (1 patient had had chronic hemorrhoidal bleeding and was already closely being followed in the outpatient clinic).

Patient noncompliance with clinicians’ recommendations for treatment was a minimal factor in reducing effective therapeutic yields. When patient noncompliance was discounted from the therapeutic yields, the adjusted yields were 3.7% for glucose testing (vs 3.4% when including noncompliance), 16.0% for total cholesterol level (vs 15.2%), 8.8% for triglyceride levels (vs 8.5%), 1.8% for HDL-C levels (vs 1.5%), and 4.5% for LDL-C levels (vs 4.2%). Noncompliance was not a factor for other laboratory test yields.

Multiple logistic regression analyses showed that patient covariates age and duration postinjury, as well as SCI/D severity categories, were not independently associated with treatment implementation (data not shown). Therapeutic yields for each of the motor severity categories, for instance, were not dissimilar (eg, for the lipid panel in aggregate: 17.1%, C1-T1 motor complete; 18.6%, C1-T1 motor incomplete; 16.1%, T2-L5 motor complete; 21.4%, T2-L5 motor incomplete). Moreover, exogenous (nonpatient) factors, such as the supervising SCI/D physician (ie, there were 5 independent supervising physicians overseeing care), did not affect diagnostic and therapeutic yields (χ² test, P=.606 and P=.110, respectively). This was despite statistically significant differences in the rates of ordering follow-up tests (χ² test, P=.045) and specialist consultation (χ² test, P<.001) among the providers. Although there was a slight tendency for case yields to be different among the physicians (χ² test, P=.089), there remained no difference in diagnostic or therapeutic yields among the different physicians even when only examining cases (χ² test, P=.605). Discounting patient noncompliance also did not result in a therapeutic yield difference among providers (χ² test, P=.116).

Back to Article Outline

Discussion 

Multiple preventive health care task forces and medical associations, supported by copious amounts of high-grade evidence in the non-SCI/D literature, have advocated against the routine laboratory testing of asymptomatic adults during periodic PHEs, except for lipid tests¹, ², ³, ⁶, ⁷ and less strongly, glucose levels.⁹, ¹² Our study, which is retrospective, exploratory, and preliminary in nature, portends that similar recommendations, if confirmed by future prospective research, may be appropriate for the SCI/D patient population as well. Our data revealed that of a total of 5865 routine laboratory tests performed in 350 SCI/D patients as part of their annual evaluations, 11.2% of the test results were outside of their respective reference ranges, yet the overall diagnostic yield was only 0.7% and the overall therapeutic yield was 2.1%. Eighty percent of the new diagnoses and almost 93% of the therapeutic interventions were incited by abnormalities in glucose or lipid testing.

The yields observed in our study were consistent with the previous literature for the non-SCI/D ambulatory population.¹¹, ³⁶, ³⁷ In the non-SCI/D literature, Boland et al,³⁶ for instance, reported abnormal laboratory value rates and therapeutic yields of 21.5% and 9.2%, respectively, for lipid screening, 18.0% and 2.2% for chemistry panels, and 6.6% and 1.8% for the CBC. A subsequent prospective series¹¹ of 531 patients also undergoing comprehensive ambulatory evaluations revealed therapeutic yields of 16.5% for lipids, 2.8% for chemistry panels, and 0.9% for the CBC. In a prospective study of 695 ambulatory adult outpatients making their initial clinic visit, Ruttimann et al³⁷ reported an abnormal laboratory test result rate of 6.4% for the CBC, with an overall therapeutic yield of only 0.2%. Ultimately, of course, there are no absolute values that categorically define what rates of diagnostic or therapeutic yields are significant enough to support routine laboratory surveillance. Prospective longitudinal research will be required to determine whether any of the laboratory tests included in periodic PHE programs for cohorts of patients with SCI/D results in better overall long-term health.

Given the health status differences between the general ambulatory population and the population with SCI/D, the reason why diagnostic and therapeutic yield findings in our study did not differ significantly from those published for the general population is unclear. Our multivariate analysis, however, did reveal that age, and not duration postinjury, was independently predictive of laboratory test results and predominantly constituted the explicable variability for most laboratory tests, implicating that guidelines for laboratory screening of patients with SCI/D might more closely resemble those already published for the general population. Although SCI/D severity categories also showed independent associations with numerous laboratory tests, the laboratory test variability explained by SCI/D severity category was less than age. Future longitudinal outcomes-based research, however, may ultimately identify subpopulations of patients with SCI/D who may be better served by regular laboratory screening.

It is, nonetheless, unclear why the diagnostic and therapeutic yields for many of the tests in our study and in others were so low, given the high rate of laboratory abnormalities. For instance, our study revealed 28 elevated alkaline phosphatase levels as a result of PHE screening. This resulted in only 1 follow-up test, no new diagnoses, and no new therapeutic interventions. Because this was a retrospective study, it is difficult to ascertain whether these abnormal laboratory test results were simply missed or deliberately not pursued further. If one considers, however, that the differential diagnosis for an elevated alkaline phosphatase is rather extensive (eg, healing fractures, osteomalacia, Paget’s disease, heterotopic ossification, Rickets, biliary obstruction, hepatitis, liver nodules, metastatic liver disease, adverse reactions to drugs, nonpathologic elevation), it then follows that further investigation, in the absence of clear signs of a pathologic process, would require numerous diagnostic maneuvers, most of which would likely be time consuming yet unfruitful for patients. It is thus not unreasonable that practitioners and their patients would opt not to blindly begin such a workup. In the future, prospective research will be required to discern whether higher diagnostic and therapeutic yields can, in fact, be effected by higher rates of blind follow-up or whether these extra tests primarily lead only to increased patient anxiety and utilization of clinical resources without improving long-term outcomes.

Besides longitudinal outcomes-based research, other criteria have emerged in the medical literature for assessing the appropriateness of laboratory screening tests. Two such criteria in the literature can be surmised as follows: (1) the incidence or pretest probabilities of a condition or disease should be sufficiently high enough to justify screening, with clinical efforts focused on the leading causes of morbidity and mortality (provided there are effective interventions), and (2) there should exist an asymptomatic period during which early detection and treatment of a disease state favorably impacts morbidity or mortality, in comparison to initiating therapy after symptom onset.², ³⁸, ³⁹ Although the benefits of early detection and treatment of dyslipidemia and hyperglycemic conditions are well accepted for all populations, even if routine screening for the latter is not uniformly endorsed, it is less clear what the long-term impacts are for interventions after the detection of other laboratory abnormalities. There is, as yet, no high-level evidence (as defined by the USPSTF)⁴⁰ in the SCI/D literature, for instance, that interventions for incidentally discovered low sodium, albumin, or hematocrit levels are effective in improving morbidity or mortality.

Given these formidable criteria, it is unclear why laboratory tests of questionable preventive health benefit are nevertheless so frequently used in practice in PHE programs.⁸, ⁴¹ Some⁴², ⁴³, ⁴⁴ have argued that periodic laboratory testing may support the needs of patients and health care providers beyond that which can be observed or quantified by “prevention experts” and that there may be some level of distrust among practicing clinicians regarding the ethical motivations behind preventive health guidelines. Based on focus group work, Beaulieu et al reported that, “[health care providers felt that] ‘it was easier to live with not following guidelines than with missing a diagnosis’ [and that] ‘preventive medicine experts frequently place economic considerations in front of scientific ones.’”^42(p521-2) For patients with SCI/D, there is a particular concern that disease states may be missed without routine surveillance in patients who have sensory deficits. For instance, for patients with complete tetraplegia, abdominal pathology (eg, cholecystitis) may not declare itself by classic symptomatology. Additionally, some practitioners may feel more comfortable to have routine “baseline” values available on record in the event that patients become acutely ill. Our data, nonetheless, do not suggest that periodic routine laboratory testing in a population with significant sensory loss necessarily leads to the frequent discovery of pathology inciting intervention, save for lipid and glucose tests. Moreover, there is no evidence at present, outside of anecdotal reports, that having baseline data improves long-term health care outcomes for general ambulatory patients or for patients with SCI/D.

The collection of periodic baseline laboratory data, as well as routine laboratory screening, is not necessarily without risk. It cannot be overemphasized that preventive health practices of questionable merit, despite all good intention and a veneer of innocuousness, can lead to unforeseen consequences and costs not restricted to purely economic domains. There is the potential, for instance, of unnecessary patient anxiety because of false-positive tests, as well as health risks because of invasive follow-up tests and unnecessary treatment changes. Moreover, there are potentially significant patient health care opportunity costs related to activities forgone by health care providers in order to provide, evaluate, and follow-up on periodic laboratory testing.⁴⁵ An undue emphasis on laboratory evaluations during the PHE, for instance, may limit the time and resources available for patient education and the completion of thorough clinical, functional, or psychosocial evaluations. Interestingly, of those problems uncovered by general practitioners during PHEs, a large majority are reported to be social rather than medical.⁴⁵ Recently, a survey of veterans with SCI/D showed that the 2 principal motivations patients had for participating in the PHE were (1) to get medications and supplies refilled and (2) to talk to the doctor about issues related to muscle strength and weakness, bladder care, chronic pain, digestion and bowel care, and equipment problems.⁴⁶ Future research should focus on clarifying and addressing which types of services truly meet the needs of SCI/D patients and have a significant impact on health care outcomes. Augmenting educational and counseling services and further addressing psychosocial, functional, and equipment needs, for instance, in place of services with little evidence-based support, may potentially go a long way in increasing patient satisfaction and improving patient care.

The annual PHE currently mandated to be offered to all people with SCI/D within the VA system has by no means been a static program.²⁰, ⁴⁷ The program began with the laudatory vision of improving the health care of veterans with SCI/D via comprehensive preventive care and has been evolving with respect to its specifics ever since. In addition to laboratory tests, the annual PHE includes a variety of other components (eg, immunizations, skin assessment, assessment of change in function and activities of daily living, urinary tract morphology and function assessment, review of equipment and supplies, psychosocial and vocational assessments, substance-abuse screening and counseling).⁴⁷ In the future, periodic reassessments of the tests offered and services rendered as part of the VA PHE program for people with SCI/D should be pursued to reflect state-of-the-art practices in preventive medicine as new data emerge.

Study Limitations 

Our laboratory data were cross-sectional, limiting inferences about the true longitudinal associations between laboratory results and patient covariates. Additionally, this study was limited to adult community-dwelling veteran males with SCI/D, who also happened to be predominantly white, reflecting the demographics of our referral population. Results are not necessarily generalizable to more ethnically diverse populations, non−community-dwelling adults, children, or women.

Although the electronic records were carefully reviewed, there exists a significant potential for the underestimation of the true rate of educational and other interventions because some of these interventions may have escaped our detection during data abstraction. Furthermore, because this was a retrospective study performed at a busy clinical service, documentation for education and counseling could not be expected to be complete always.

Back to Article Outline

Conclusions 

Although the comprehensive annual PHE program represents a significant advance in the health care of veterans with SCI/D in the United States, our data suggest that some of the laboratory components performed routinely as part of the PHE may be of limited utility. Many diagnostic and therapeutic yields from selected laboratory components of the PHE were low, consistent with reports in the general ambulatory population. Further data collection, particularly prospective longitudinal data, is warranted to potentially refine the selection of these tests and appropriately qualify, if necessary, the intervals for tests performed as part of the periodic PHE program. Such data would be valuable in further improving the impact of PHE programs on the morbidity and mortality of veterans and other people with SCI/D.

Supplier

Back to Article Outline

Acknowledgment 

We thank E. Francis Cook, ScD, Harvard School of Public Health, for his many helpful suggestions on the manuscript.

Back to Article Outline

References 

1. U.S. Preventive Services Task Force . Guide to clinical preventive services: an assessment of the effectiveness of 169 interventions: Report of the U.S. Preventive Services Task Force . Baltimore: Williams & Wilkins; 1989;
   • View In Article
2. U.S. Preventive Services Task Force . The periodic health examination (age-specific charts) . Am Fam Physician . 1990;41:189–204
   • View In Article
   • MEDLINE
3. U.S. Preventive Services Task Force . In: Guide to clinical preventive services . 2nd ed.. Baltimore: Williams & Wilkins; 1996;
   • View In Article
4. Medical Practice Committee American College of Physicians . Periodic health examination (a guide for designing individualized preventive health care in the asymptomatic patient) . Ann Intern Med . 1981;95:729–732
   • View In Article
   • MEDLINE
5. Council on Scientific Affairs . Medical evaluation of healthy persons . JAMA . 1983;249:1626–1632
   • View In Article
   • MEDLINE
6. Canadian Task Force on the Periodic Health Examination . The periodic health examination . CMAJ . 1979;121:1193–1254
   • View In Article
7. Canadian Task Force on the Periodic Health Examination . The periodic health examination (2, 1985 Update) . CMAJ . 1986;134:724–727
   • View In Article
   • MEDLINE
8. van Walraven C , Goel V , Austin P . Why are investigations not recommended by practice guidelines ordered at the periodic health examination? . J Eval Clin Pract . 2000;6:215–224
   • View In Article
   • MEDLINE
   • CrossRef
9. Silverstein MD , Boland BJ . Conceptual framework for evaluating laboratory tests (case-finding in ambulatory patients) . Clin Chem . 1994;40:1621–1627
   • View In Article
   • MEDLINE
10. Ruttimann S , Dreifuss M , Clemencon D , di Gallo A , Dubach UC . Multiple biochemical blood testing as a case-finding tool in ambulatory medical patients . Am J Med . 1993;94:141–148
    • View In Article
    • Abstract
    • Full-Text PDF (1022 KB)
    • CrossRef
11. Boland BJ , Wollan PC , Silverstein MD . Yield of laboratory tests for case-finding in the ambulatory general medical examination . Am J Med . 1996;101:142–152
    • View In Article
    • Abstract
    • Full-Text PDF (1121 KB)
    • CrossRef
12. Grundy SM , Greenland P , Herd A , et al.   Cardiovascular and risk factor evaluation of healthy American adults. A statement for physicians by an Ad Hoc Committee appointed by the Steering Committee, American Heart Association . Circulation . 1987;75:1340A–1362A
    • View In Article
    • MEDLINE
13. Shapiro MF , Greenfield S . The complete blood count and leukocyte differential count. An approach to their rational application . Ann Intern Med . 1987;106:65–74
    • View In Article
    • MEDLINE
14. The South-East London Screening Study Group . A controlled trial of multiphasic screening in middle-age (results of the South-East London Screening Study) . Int J Epidemiol . 1977;6:357–363
    • View In Article
    • MEDLINE
15. Kim DE , Berlowitz DR . The limited value of routine laboratory assessments in severely impaired nursing home residents . JAMA . 1994;272:1447–1452
    • View In Article
    • MEDLINE
16. McKinley WO , Jackson AB , Cardenas DD , DeVivo MJ . Long-term medical complications after traumatic spinal cord injury (a regional model systems analysis) . Arch Phys Med Rehabil . 1999;80:1402–1410
    • View In Article
    • Abstract
    • Full-Text PDF (1195 KB)
    • CrossRef
17. Bauman WA , Adkins RH , Spungen AM , et al.   Is immobilization associated with an abnormal lipoprotein profile? Observations from a diverse cohort . Spinal Cord . 1999;37:485–493
    • View In Article
    • MEDLINE
18. Fong TL , Adkins RH , Govindarajan S , Post S , Waters RL . Prevalence of hepatitis C infection in a large urban hospital-based sample of individuals with spinal cord injury . Arch Phys Med Rehabil . 2002;83:1620–1623
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (54 KB)
    • CrossRef
19. Knox EG . Multiphasic screening . Lancet . 1974;2:1434–1436
    • View In Article
    • MEDLINE
20. Veterans Health Administration Handbook 1176.1. Spinal cord injury and disorders system of care procedures . Washington (DC): Department of Veterans Affairs; 2002;
    • View In Article
21. American Spinal Injury Association . International standards for classification of spinal cord injury, revised 2002 . Chicago: ASIA; 2002;
    • View In Article
22. Medical Research Council . Aids to the investigation of peripheral nerve injury . London: H.M. Stationery Office; 1942; War memorandum No. 7
    • View In Article
23. Chambers JM , Cleveland WS , Kleiner B , Tukey PA . Graphical methods for data analysis . Belmont: Duxbury Pr; 1983;
    • View In Article
24. Cleveland WS , Devlin SJ . Locally weighted regression (an approach to regression analysis by local fitting) . J Am Stat Assoc . 1988;83:596–610
    • View In Article
25. Silverman BW . Kernel density estimation using the fast Fourier transform . Appl Stat . 1982;31:93–99
    • View In Article
26. Mohler JL , Barton SD , Blouin RA , Cowen DL , Flanigan RC . The evaluation of creatinine clearance in spinal cord injury patients . J Urol . 1986;136:366–369
    • View In Article
    • MEDLINE
27. Guralnik JM , Eisenstaedt RS , Ferrucci L , Klein HG , Woodman RC . Prevalence of anemia in persons 65 years and older in the United States (evidence for a high rate of unexplained anemia) . Blood . 2004;104:2263–2268
    • View In Article
    • MEDLINE
    • CrossRef
28. Cockcroft DW , Gault MH . Prediction of creatinine clearance from serum creatinine . Nephron . 1976;16:31–41
    • View In Article
    • MEDLINE
    • CrossRef
29. Ferrara A , Barrett-Connor E , Shan J . Total, LDL, and HDL cholesterol decrease with age in older men and women. The Rancho Bernardo Study 1984-1994 . Circulation . 1997;96:37–43
    • View In Article
    • MEDLINE
30. McPherson K , Healy MJ , Flynn FV , Piper KA , Garcia-Webb P . The effect of age, sex and other factors on blood chemistry in health . Clin Chim Acta . 1978;84:373–397
    • View In Article
    • MEDLINE
    • CrossRef
31. Bauman WA , Adkins RH , Spungen AM , Kemp BJ , Waters RL . The effect of residual neurological deficit on serum lipoproteins in individuals with chronic cord injury . Spinal Cord . 1998;36:13–17
    • View In Article
    • MEDLINE
32. Sica DA , Midha M , Zawada E , Stacy W , Hussey R . Hyponatremia in spinal cord injury . J Am Paraplegia Soc . 1990;13:78–83
    • View In Article
    • MEDLINE
33. Perkash A , Brown M . Anemia in patients with traumatic spinal cord injury . J Am Paraplegia Soc . 1986;9:10–15
    • View In Article
    • MEDLINE
34. Duckworth WC , Solomon SS , Jallepalli P , Heckemeyer C , Finnern J , Powers A . Glucose intolerance due to insulin resistance in patients with spinal cord injuries . Diabetes . 1980;29:906–910
    • View In Article
    • MEDLINE
35. MacDiarmid SA , McIntyre WJ , Bailey RR , Turner JG , Arnolds EP . Monitoring of renal function in patients with spinal cord injury . BJU Int . 2000;85:1014–1018
    • View In Article
    • MEDLINE
    • CrossRef
36. Boland BJ , Wollan PC , Silverstein MD . Review of systems, physical examination, and routine tests for case-finding in ambulatory patients . Am J Med Sci . 1995;309:194–200
    • View In Article
    • MEDLINE
    • CrossRef
37. Ruttimann S , Clemencon D , Dubach UC . Usefulness of complete blood counts as a case-finding tool in medical outpatients . Ann Intern Med . 1992;116:44–50
    • View In Article
    • MEDLINE
38. Frame PS , Carlson SJ . A critical review of periodic health screening using specific screening criteria. Part 4: selected miscellaneous diseases . J Fam Pract . 1975;2:283–289
    • View In Article
    • MEDLINE
39. Frame PS . A critical review of adult health maintenance: Part 4. Prevention of metabolic, behavioral, and miscellaneous conditions . [published erratum in: J Fam Pract 1986;23:537] J Fam Pract . 1986;23:29–39
    • View In Article
    • MEDLINE
40. Methods Work Group. Third U.S. Preventive Services Task Force . Current methods of the U.S. Preventive Services Task Force (a review of the process) . Am J Prev Med . 2001;20(3 Suppl):21–35
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (156 KB)
    • CrossRef
41. Romm FJ . “Routine” chemistry testing . Fam Med . 1986;18:230–232
    • View In Article
    • MEDLINE
42. Beaulieu MD , Hudon E , Roberge D , Pineault R , Forte D , Legare J . Practice guidelines for clinical prevention (do patients, physicians and experts share common ground?) . CMAJ . 1999;161:519–523
    • View In Article
    • MEDLINE
43. Oboler SK , Prochazka AV , Gonzales R , Xu S , Anderson RJ . Public expectations and attitudes for annual physical examinations and testing . Ann Intern Med . 2002;136:652–659
    • View In Article
44. Han PKJ . Historical changes in the objectives of the periodic health examination . Ann Intern Med . 1997;127:910–917
    • View In Article
    • MEDLINE
45. Shackley P , Donald S . Prevention in primary care (the annual assessment of elderly people) . Health Policy . 1993;25:51–62
    • View In Article
    • MEDLINE
    • CrossRef
46. Collins EG , Langbein WE , Smith B , Hendricks R , Hammond M , Weaver F . Patients’ perspective on the comprehensive preventive health evaluation in veterans with spinal cord injury . Spinal Cord . 2005;43:366–374
    • View In Article
    • MEDLINE
    • CrossRef
47. Veterans Health Administration . Spinal cord injury service . Veterans Health Administration Manual M-2: Clinical affairs. Pt XXIV . Washington (DC): Department of Veterans Affairs; 1994;
    • View In Article

• a Version 8.2; Stata Corp, 4905 Lakeway Dr, College Station, TX 77845.