Search

Background: Interpretation of serial data in monitoring of patients is usually performed by use of the 'reference change value' (RCV). While this tool for interpretation of measured differences is simple and clear, there are a number of drawbacks attached to the uncritical use of this concept. It is a dichotomised interpretation of continuous data using a fixed probability without any counter hypothesis. Therefore, a tool for better understanding and interpretation of measured differences in monitoring is needed. Theory: The concept of sensitivity, specificity, likelihood ratios and odds used for diagnostic test evaluations is applied to monitoring by substituting measured concentrations with measured differences. Thus, two frequency distributions of differences are assumed, one for a stable, steady-state, situation and one for a certain change. Values exceeding a measured difference will thus represent 'false change' for the stable and 'true change' for the change and the ratio between these will define the likelihood. By making the hypothesis of change variable and equal to the actual difference, the distribution corresponding to the true changes for the measured difference varies with this. Consequently, the likelihood ratio for change increases with growing measured difference and when used together with the pre-test odds or pre-test probability, the post-test odds and post-test probability, related to the clinical situation, can be calculated. Results: One example is acute intermittent porphyria, where increasing excretion of porphobilinogen is characteristic for an attack. The within-subject biological variation is estimated to 25%, which for two measurements gives a variation of 35% for measured differences. Three pre-test probabilities are assumed and illustrate that post-test odds and probability depends on both the pre-test probability and the measured difference. A second example is monitoring women in a follow-up after treatment of breast cancer, using the tumour marker CA 15-3. The within-subject biological variation is estimated to 14.9% and for differences 21% (2½ X 14.9 due to two measurements). Here, the monitoring is totally scheduled and the frequency of progression depends on the time after treatment. Thus, the pre-test probability varies with time so that a certain measured difference with a given likelihood ratio will result in varying post-test odds depending on time. Conclusions: The concept presented here expands the earlier concept of RCVs by making it possible to have an estimate of the post-test odds for a certain difference to occur based on likelihood ratios and pre-test odds. [ABSTRACT FROM AUTHOR]

Background: In modern medicine many laboratory measurements are used as a surrogate for clinical outcome. There is therefore a need for models to quantify this approach. The target plot presented here is based on the difference plot. The difference between two measurements is plotted against the initial measurement, of which the first measurement is the index measurement (Mindex) and the second the outcome measurement (Moutcome). A vertical line separates normal vs. increased Mindex concentrations according to a selected target concentration, and a horizontal line describes the situation of no change. A target line (Mtarget) with a slope equal to −1 and indicating the selected target concentration is drawn through the crossing point. The six outcome areas thus refer to the combined effect of Mindex and Moutcome. A target plot taking the biological and analytical variation in consideration is considered. Materials and model: The target plot addresses sources of outcome categories, and calculates the percentage in each of these categories according to a defined target value. Hemoglobin A1c (HbA1c) in diabetes in patients from Vejle County, Denmark was taken as an example. Results: Different strategies for target setting for HbA1c were investigated: the Danish Diabetes Association (DDA) target of 5.84% HbA1c, the target from the American Diabetes Association (ADA) of 7.0% HbA1c, and 6.62% HbA1c, which is the 99.9 centile of the Danish national reference interval. All three strategies can be validated from the target plot for differences in surrogate clinical outcome in spite of identical patient material. Use of the ADA target (the highest nominal value) reveals the lowest percentage of index values above the target, and of those the highest percentage outcome values of the surrogate measure reached the target. Conclusion: The target plot allows detailed stratification of outcome based on surrogate parameters and assessment of risk of disease based on the selected target. The need to standardize outcome targets in diabetes to allow comparison of quality of treatment is obvious. [ABSTRACT FROM AUTHOR]

Background: Hemoglobin A1c (HbA1c) has been used in controlled trials for the last 10 years but has never been evaluated in clinical practice as an effective parameter for clinical outcome. We investigated the trend for glycemic control over 11 years in one county of 350,000 citizens. Methods: We studied 226,382 HbA1c-DCCT-aligned from 39,455 patients in whom routine monitoring for diabetes was initiated in 1993, 1996, or 2001. The trend in glycemic control was investigated in groups by probit plots, and in individual patients by target plots. Results: From 1993 to 2001, the number of HbA1c measurements increased three-fold. The number of new monitoring series increased from 0.22% to 0.27% of the county population, and the number of patients monitored using HbA1c as a proxy for diabetes increased from 0.5% to 1.5%. A proportional reduction in high HbA1c concentrations of 5% was identified in the 1993 group, compared to 15% in the 1996 group, and 20% in the 2001 group. The percentage of patients with diabetic first HbA1c experiencing normalization increased from 8% to 30% for males and from 9% to 24% for females (1993–2001). The percentage of HbA1c concentrations that were not normalized decreased from 78% to 53% for males and from 83% to 59% for females (1993–2001). The median HbA1c at initiation of monitoring decreased from 8.7% in 1993 to 7.5% in 2001 (p<10?5). The number of normal first HbA1c results in monitoring series increased from 7% to 17% for males and from 8% to 22% for females. Up to 10% of subjects developed diabetic concentrations during monitoring. Conclusion: On average, patients with diabetic first HbA1c concentrations (?6.62%) showed an improvement in glycemic control from 5% in 1993 to 20% in 2001. High concentrations were easiest to reduce. In patients with originally diabetic HbA1c levels, 66% on average showed improved glycemic control in the 2001 series compared to 50% in the 1993 series. An average of 6% (1993) vs. 9% (2001) with originally normal HbA1c levels showed an upward trend inHbA1c levels. Median HbA1c at initiation of monitoring decreased from 8.7% in 1993 to 7.5% in 2001 (p<10?5). The incidence of new cases was constant. [ABSTRACT FROM AUTHOR]

Background: Hemoglobin A1c (HbA1c) is a proxy measure for glycemic control in diabetes. We investigated the trend for glycemic control in patients from three Danish counties using HbA1c measurements. Methods: We studied 2454 patients from a population of 807,000 inhabitants for whom routine monitoring of diabetes using HbA1c-DCCT aligned was initiated in 2001. We estimated the incidence of monitored patients in the population. The progress in patients with originally diabetic HbA1c levels was investigated by cumulative probability plots, and the individual trend in clinical outcome was investigated by a modified difference plot. Results: The age-standardized incidence of monitored patients was <0.5% in all regions. Patients with diabetic first HbA1c concentrations (≥6.62% HbA1c) showed on average 15% improved glycemic control in the first year. Further improvement was limited. The overall percentage above the treatment target (≥6.62% HbA1c) was 51% in 2003 compared to 59% in 2001, and the percentage with poor glycemic control (≥10.0% HbA1c) was reduced from 19% to 4%. Of patients with originally diabetic HbA1c levels, 15% showed progress in glycemic control, and 28% reached treatment targets. In patients with originally normal HbA1c, 75% showed an upward trend in HbA1c levels, which reached diabetic concentrations in 17%. Conclusion: Patients with diabetic first HbA1c concentrations (≥6.62% HbA1c) showed on average 15% improved glycemic control in the first year. Further improvement was limited. In individual patients, 75% with originally diabetic HbA1c levels showed improved glycemic control after 3 years, while 78% withoriginally normal concentrations showed an upwardtrend in HbA1c levels. [ABSTRACT FROM AUTHOR]

Clinical activity indices are essential instruments in monitoring inflammatory bowel diseases such as Crohn's disease (CD) and ulcerative colitis (UC). To subclassify components of disease indices in CD and UC, investigate technical noise in estimation of the indices, establish a signal-to-noise ratio (SNR), evaluate correlation between indices and calculate the reference change value (RCV) for selected biochemical variables in individual cases, 50 patients with CD and 49 patients with UC were included in the study. Qualitative index variables were assessed for scoring errors. The standard deviation (SD) was estimated according to a rectangular model, while SD in biochemical variable scoring was estimated according to a Gaussian model; a combined SD was also calculated. These values were investigated for their individual contribution to variation. The 95% CI of an index value was based on ±1.96×SD combined and a change in separate biochemical variables was calculated as RCV 1.96×√2×SD combined . Correlation between different disease activity indices was assessed for unexplained variation. The Crohn's disease activity index (CDAI) had the highest variation compared to the van Hees (Hees) and the Harvey-Bradshaw index (HBI) in CD, but it also had the best SNR, whereas HBI had the lowest. In UC the clinical activity index (CAI) showed the highest variance, but the best SNR compared to Seo's activity index (AI). The 95% CI of the CDAI discriminatory activity sum of 150 in individual cases was 105–195, whereas the 95% interval for a change was ±62.4. Self-reported wellness contributed 40% to total variance in the CDAI. Factors of clinical importance increased errors in estimates and variance of the indices. Poor correlation was obtained between activity indices, with up to 70% unexplained variance. The SD combined for estimated errors was as high as 23 points, with the best SNR being approximately 20. Index factors increase the sensitivity of SNRs to errors and lower the disease specificity. Sensitivity optimisation may be achieved by standardisation of the variables and their use. [ABSTRACT FROM AUTHOR]

Gestational diabetes mellitus (GDM) is defined as carbohydrate intolerance during pregnancy. In Denmark the health service offers selective screening for GDM, i.e., admission to an oral glucose tolerance test (OGTT) after pre-screening with interview for clinical risk factors for GDM, two capillary fasting blood glucose (cFBG) measurements and a urine test for glucosuria. The aim of the present study was to investigate the power of the pre-screening to identify GDM and the screening to predict adverse clinical outcome. A retrospective investigation of pregnant women undergoing screening during 1998 at Vejle County Hospital, Denmark was undertaken. The two most frequent pre-screening criteria for OGTT were bodymass index (BMI) &ges;27 kg/m² and age &ges;35 years. The highest odds ratio (OR) of 9.07 (95% CI: 2.60 to 63.70) for GDM had glucosuria and the lowest (zero) had cFBG. The frequency of complicated delivery was similar in GDM (58%) compared to non-GDM (56%). The best predictor of complicated delivery was a BMI with OR=1.50 (95% CI: 0.87 to 2.60) for BMI &gts;27 kg/m² vs. <27 kg/m². The best predictor of adverse neonatal outcome was a capillary blood glucose 120 min after glucose load (cBG120 min) &ges;9.0 mmol/l (OR=3.18, 95% CI: 1.14 to 8.89). The intermediary endpoint GDM was not superior for predicting adverse maternal and neonatal outcome. The cumulative probability distribution of cBG120 min after a 75 g glucose load was not homogeneously distributed in groups stratified according to maternal and foetal outcome. A changed slope was seen after cBG120 min 9.0 mmol/l. Screening cFBG of 4.1 mmol/l was unable to predict GDM and adverse outcome. Glucosuria was too rare to be effective as a screening tool. Pre-screening did not identify GDM. The best predictor of complicated delivery was a high BMI. The best predictor of foetal adverse outcome was cBG120 min ≥9.0 mnol/l after a 75 g glucose load. Identical fraction complications were present in GDM and non-GDM. A refinement of the screening procedure is highly needed, and this has been initiated in Denmark. [ABSTRACT FROM AUTHOR]

The reference interval is probably the most widely used decision-making tool in clinical practice, with a modern use aiming at identifying wellness during health check and screening. Its use as a diagnostic tool is much less recognised and may be obsolete. The present study investigates the consequences of the new practice for the interpretation of prospective value, negative vs. positive, the probability of con- firming wellness, and number of false results based on selected strategy for reference interval establishment. Calculations assumed normalised Gaussian-distributed reference intervals with analytical variation set to zero and absolute accuracy. Also assumed is the independency of tests. Probability for no values out- side reference intervals in healthy subjects was calculated from the formula Pno outside = (1 - Psingle) and according to the formula for repeated testing: Pone outside = n × Psingle (1 - Psingle)n-1 etc. Here n is the number of tests performed and Psingle is the probability of one result outside reference limits with the general formula pi outside n- i = k × Psinglei (1 - Psingle)n-i , with k being the binominal coefficient and i the number outside the reference intervals. Use of the 99.9 centile for health checks will increase the probability for no false from 60% to 99% for 10 tests, and from 46% to 98% for 15 tests. The probability for one false-positive result in 10 tests in a panel can be reduced from 32% to 1% if the 99.9% centile is substituted for the 95% centile. For two in 10 tests, the probability can be reduced from 8% to below 0.1%. In both cases, selection of the 99.9% centile improves the diagnostic accuracy. Reference intervals are needed as a "true" negative reference for absence of disease, and should cover the 99.9% centile of the reference distribution of an analyte to avoid false positives. For this new use, it is critical that reference persons are absolutely normal without clinical, genetic and biochemical signs of the condition being investigated. However, reference intervals cannot substitute clinical decision limits for diagnosis and medical intervention. [ABSTRACT FROM AUTHOR]

Reference intervals are recommended for naturally occurring quantities and required in the evaluation of new components in order to provide clinically useful information. The aim of the present study is to present a method for selecting reference individuals for the determination of fasting venous plasma glucose (f-vPG) reference intervals and ways to determine if disease groups can share reference intervals with an ideal reference population. Reference subjects were randomly selected, eligibility was judged according to predetermined inclusion and exclusion criteria. Using the literature we selected risk indicators for diabetes mellitus (DM) and used these indicators to rule out high-risk individuals in order to obtain a reference distribution of f-vPG determined using individuals with low risk of DM. The distributions of f-vPG in the high-risk individuals was compared with that determined for the low-risk group. We then estimated the ability of the high-risk individuals to share the reference interval of the low- risk individuals, and calculated the fraction that was outside this interval. Distributions were also investigated for linearity in the cumulated frequency rankit distribution of In-values. The allowable difference between two reference limits could not exceed 0.375 times the population biological variation. Most risk indicators were powerful predictors of high f-vPG values. Subgroups with these risk indica- tors should not be included in the homogeneous In-normally distributed reference distribution. Distributions of f-vPG concentrations in individuals with risk factors were not homogeneous and varying percentages of individuals were outside the reference distribution, having f-vPG greater than 7.0 mmol/l. We conclude that randomisation is only useful to recruit candidate reference subjects. To rule out subjects according to clinical risk factors for diabetes, it is necessary to identify a reference population with low risk of exhibiting increased f-vPG concentrations. This method may be used to validate a reference interval for a particular analyte with respect to an investigated disease, and to stratify risk factors of importance. [ABSTRACT FROM AUTHOR]

A well-known transformation from the bell-shaped Gaussian (normal) curve to a straight line in the rankit plot is investigated, and a tool for evaluation of the distribution of reference groups is presented. It is based on the confidence intervals for percentiles of the calculated Gaussian distribution and the percentage of cumulative points exceeding these limits. The process is to rank the reference values and plot the cumulative frequency points in a rankit plot with a logarithmic (In = log8) transformed abscissa. If the distribution is close to In-Gaussian the cumulative frequency points will fit to the straight line describing the calculated In-Gaussian distribution. The quality of the fit is evaluated by adding confidence intervals (Cl) to each point on the line and calculating the percentage of points outside the hyperbola-like Cl-curves. The assumption was that the 95% confidence curves for percentiles would show 5% of points outside these limits. However, computer simulations disclosed that approximate 10% of the series would have 5% or more points outside the limits. This is a conservative validation, which is more demanding than the Kolmogorov-Smirnov test. The graphical presentation, however, makes it easy to disclose deviations from In-Gaussianity, and to make other interpretations of the distributions, e.g., comparison to non-Gaussian distributions in the same plot, where the cumulative frequency percentage can be read from the ordinate. A long list of examples of In-Gaussian distributions of subgroups of reference values from healthy individuals is presented. In addition, distributions of values from well-defined diseased individuals may show up as In-Gaussian. It is evident from the examples that the rankit transformation and simple graphical evaluation for non-Gaussianity is a useful tool for the description of sub-groups. [ABSTRACT FROM AUTHOR]

The article presents the author's views on the significance of laboratory tests. The need for evidence-based medical decisions has focused on quantitative data as convenient parameters for clinical handling of patients. Most laboratories furnish fast electronic test responses and can provide information on a variety of technical data on these tests. Therefore, the findings from studies of various types must be validated in clinical studies in daily routine to prove their value in patient care. Moreover, laboratory tests have become more central to modern medicine and deserve better evaluation than has been the norm in the past. In modern laboratory medicine the necessary pre-requisites for a valid evaluation are available.