SEARCH · PubMed Health

Results for “parameter estimation”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,171 records · Page 65Linked to original sources

PubMed1995

Albuminuria and 24-h ambulatory blood pressure in normoalbuminuric and microalbuminuric NIDDM patients. A longitudinal study.

OBJECTIVE: To assess the long-term relationships between 24-h ambulatory blood pressure (AMBP), urinary albumin excretion (UAE) rate, and metabolic control in non-insulin-dependent diabetes mellitus (NIDDM) patients with normo- and microalbuminuria. RESEARCH DESIGN AND METHODS: We conducted a prospective study of 23 NIDDM patients (11 with normoalbuminuria and 12 with microalbuminuria) receiving standard clinical care, including antihypertensive treatment, attending the outpatient clinic and 8 healthy control subjects. Twenty-four-hour AMBP and UAE were measured synchronously in addition to fasting plasma glucose, HbA1c, and serum creatinine at baseline and after 4.6 (4.2-5.1) years [mean (range)]. RESULTS: Baseline systolic, but not diastolic, 24-h AMBP was significantly higher in diabetic patients compared with control subjects (146/80 [16/11] vs. 133/78 [9/9] mmHg, P < 0.05), but was similar in normoalbuminuric (143/81 [11/11] mmHg) and microalbuminuric (148/80 [20/10] mmHg) patients during strict blood pressure control. The annual increase in 24-h AMBP was equivalent in diabetic patients (0.6/-0.2 [2.6/1.5] mmHg/year) and control subjects (0.7/0.2 [1.2/1.4] mmHg/year, NS) and not significantly different from zero. Overall UAE did not change in control subjects (5.6 [1.6] vs. 4.4 [1.9]) (geometric mean [antilog SD]) or in the normoalbuminuric (8.7 [1.7] vs. 11.3 [3.0] micrograms/min) and microalbuminuric (35.7 [2.1] vs. 34.5 [3.2] micrograms/min) patients. In diabetic patients, the annual change in UAE correlated significantly with the annual change in the systolic (r = 0.61, P < 0.002) and diastolic (r = 0.54, P < 0.008) 24-h AMBP. In microalbuminuric patients, only the annual increase in systolic 24-h AMBP correlated significantly with the annual change in UAE (r = 0.71, P = 0.010), whereas in the normoalbuminuric patients, only the annual increase in diastolic 24-h AMBP and the annual change in UAE were significantly correlated (r = 0.66, P = 0.026). In a stepwise multiple linear regression analysis, the annual progression in albuminuria in NIDDM patients was significantly determined by increases in systolic (parameter estimate 0.018, SE 0.006, P < 0.008) as well as in diastolic 24-h AMBP (parameter estimate 0.026, SE 0.011, P < 0.033). CONCLUSIONS: In an outpatient clinical setting, 24-h AMBP is similar in NIDDM patients with normo- and microalbuminuria. Alterations in both 24-h AMBP and UAE are on average moderate and equivalent compared with those in healthy control subjects. Although the average change in albuminuria is small, a progression in albuminuria relates to increments in both systolic and diastolic 24-h AMBP.

Albuminuria↗

PubMed2002

Estimates of parameters between direct and maternal genetic effects for weaning weight and direct genetic effects for carcass traits in crossbred cattle.

Estimates of heritabilities and genetic correlations were obtained for weaning weight records of 23,681 crossbred steers and heifers and carcass records from 4,094 crossbred steers using animal models. Carcass traits included hot carcass weight; retail product percentage; fat percentage; bone percentage; ribeye area; adjusted fat thickness; marbling score, Warner-Bratzler shear force and kidney, pelvic and heart fat percentage. Weaning weight was modeled with fixed effects of age of dam, sex, breed combination, and birth year, with calendar birth day as a covariate and random direct and maternal genetic and maternal permanent environmental effects. The models for carcass traits included fixed effects of age of dam, line, and birth year, with covariates for weaning and slaughter ages and random direct and maternal effects. Direct and maternal heritabilities for weaning weight were 0.4 +/- 0.02 and 0.19 +/- 0.02, respectively. The estimate of direct-maternal genetic correlation for weaning weight was negative (-0.18 +/- 0.08). Heritabilities for carcass traits of steers were moderate to high (0.34 to 0.60). Estimates of genetic correlations between direct genetic effects for weaning weight and carcass traits were small except with hot carcass weight (0.70), ribeye area (0.29), and adjusted fat thickness (0.26). The largest estimates of genetic correlations between maternal genetic effects for weaning weight and direct genetic effects for carcass traits were found for hot carcass weight (0.61), retail product percentage (-0.33), fat percentage (0.33), ribeye area (0.29), marbling score (0.28) and adjusted fat thickness (0.25), indicating that maternal effects for weaning weight may be correlated with genotype for propensity to fatten in steers.

Adipose Tissue↗

PubMed2005

Evaluation of autoregressive covariance structures for test-day records of Holstein cows: estimates of parameters.

The objective of this study was to compare test-day (TD) models with autoregressive covariance structures for the estimation of genetic and environmental components of variance for milk, fat and protein yields, and somatic cell score (SCS) in Holstein cows. Four models were compared: model I (CS model) was a simple TD repeatability animal model with compound symmetry covariance structure for environmental effects, model II (ARpe model) and model III (ARe model) had first-order autoregressive covariance structures for TD permanent or residual environmental effects, respectively, and model IV (305-d model) was a simple animal model using 305-d records. Data were 106,472 first-lactation TD records of 12,071 Holstein cows calving from 1996 through 2001. Likelihood ratio tests indicated that ARpe and ARe models fit the data significantly better than the CS model. The ARe model resulted in slightly smaller estimates of genetic variance and heritability than did the CS model. Estimates of residual variance were always smaller with the CS model than with the ARe model with the autoregressive covariance structure among TD residual effects. Estimates of heritability with different TD models were in the range of 0.06 to 0.11. The 305-d model resulted in estimates of heritability in the range of 0.11 to 0.36. The autoregressive covariance structure among TD residual effects may help to prevent bias in heritability estimates for milk, fat and protein yields, and SCS.

Analysis of Variance↗

PubMedDate not supplied

Importance of the vadose zone in analyses of unconfined aquifer tests.

Analytical models commonly used to interpret unconfined aquifer tests have been based on upper-boundary (water table) conditions that do not adequately address effects of time-varying drainage from the vadose zone. As a result, measured and simulated drawdown data may not agree and hydraulic parameters may be inaccurately estimated. A 72-hour aquifer test conducted in Cape Cod, Massachusetts, in a slightly heterogeneous, coarse-grained, glacial outwash deposit was found to be a good candidate for testing models with different upper-boundary conditions. In general, under the commonly invoked assumption of instantaneous drainage, measured and simulated drawdowns were found to agree with one another only at late time and early time. In the intermediate-time range, because of delayed drainage, measured drawdowns always exceeded simulated values, most noticeably in piezometers located near the water table. To reduce these discrepancies, an analytical model was developed that can fully account for time-varying drainage given that the aquifer is not strongly heterogeneous. The approach is flexible as the model, which makes use of empirical relations, does not constrain drainage to follow any particular functional relation. By this approach, measured and simulated drawdowns agree over the complete time range, and the estimated parameters are consistent with prior studies and with what is known about the aquifer geometry, stratigraphy, and composition. By properly accounting for vadose zone drainage, it was found that realistic estimates of all hydraulic parameters, including specific yield, could be obtained with or without the use of late-time data.

Geologic Sediments↗

PubMed1985

INTRAV and ORAL: BASIC interactive computer programs for estimating pharmacokinetic parameters.

Two interactive computer programs, INTRAV and ORAL, were written to permit pharmacokinetic modeling of experimental data and to obtain pertinent values based on derived estimates. Both programs utilize BASIC language and were developed on a microcomputer with graphics capability. Drug concentration in blood, plasma, or serum with time following either nonabsorptive (intravenous) or absorptive (oral or intramuscular) administration is input, and a semilogarithmic display of data appears on a cathode-ray tube (CRT). The user selects limits for various linear segments using a movable cursor. On command, coefficients and exponents for the differential equation which describes those limits is computed and a nonlinear curve is fitted through the data set. Results from statistical tests are available in output formats permitting the user to determine the goodness of the selected limits. Commonly used pharmacokinetic parameters are also computed and appear on the output. Numerous graphic output options are also available to permit comparisons between data sets and/or estimates derived from other computer programs. INTRAV and ORAL were compared with the widely used programs CSTRIP, ESTRIP, and NONLIN. Both INTRAV and ORAL gave estimates which were almost identical with CSTRIP and ESTRIP, whereas those obtained with NONLIN were very similar, although not identical.

Computers↗

PubMedDate not supplied

Identifiability of segregation parameters using estimating equations.

To eliminate the need for distributional assumptions and to reduce the computational burden associated with the method of maximum likelihood, several researchers have proposed using estimating equations techniques for segregation analysis. One concern with the application of this technique has been that the first and second order moments may not carry sufficient information for identifying all of the parameters in segregation models. It is shown that in addition to the marginal means and covariances from nuclear family data, up to the third order product moments need to be used in estimating equations for identifying all of the segregation parameters in a major gene model. A polygenic component and potentially a common family environment parameter can also be identified using up to the fourth order moments. Two weighting functions are developed to improve statistical efficiency.

Environment↗

PubMed1980

Estimating Kinetic Parameters when the Amount of Enzyme Added to an Assay Is Not a Controlled Variable: NITROGENASE ACTIVITY OF INTACT LEGUMES .

Reliable estimates of Michaelis constants (K(m)) and inhibitor constants may be obtained, in the absence of control over the amount of enzyme being added to any assay system, provided the following constraints are met. Michaelis-Menten kinetics are obeyed. Two rate measurements must be made with the same sample of enzyme: at low and high substrate concentration for determining K(m) or minus and plus an inhibitor for determining inhibitor constants. The Michaelis constant may be calculated from the equation [Formula: see text] Inhibitor constants are derived graphically from Lineweaver-Burk or Dixon plots, once the K(m) has been calculated. The above technique has been applied to study of the acetylene-reducing ability of intact legume plants. The apparent K(m) for acetylene reduction by nitrogenase in legume nodules is approximately 1/100 atmosphere in the absence of nitrogen and approximately 1/40 atmosphere in its presence.

Journal Article↗

PubMed2006

Quantitative analysis of 11C-verapamil transfer at the human blood-brain barrier for evaluation of P-glycoprotein function.

UNLABELLED: P-glycoprotein in the blood-brain barrier (BBB) has been found to be associated with several types of neurologic damage. (11)C-Verapamil has been used for in vivo imaging of P-glycoprotein function in the BBB by PET, but metabolites in plasma complicate the quantitative analysis of human studies. In this study, we validated the quantification methods of (11)C-verapamil transfer from plasma to the brain in humans. METHODS: The transfer rate constant from plasma to the brain, K(1), was estimated by nonlinear least squares (NLS) with a 2-input compartment model, including the permeation of the main metabolite in plasma at the BBB, and with a 1-input compartment model using only 15-min data that contained little metabolite in plasma. K(1) was also estimated by graphical analysis of an integration plot that uses only early-time data, before the appearance of metabolites, and the estimated K(1) was compared with that obtained by the NLS method. In the simulation study, the reliability of parameter estimates in the graphical analysis method was investigated for various values of rate constants, time ranges of parameter estimations, and noise levels. RESULTS: (11)C-Verapamil in plasma gradually converted to its metabolites, and about 45% of the radioactivity in the plasma specimen was associated with (11)C-verapamil metabolites at 30 min after injection. Although K(1) estimated from graphical analysis was slightly smaller than that by NLS, there was strong correlation among the K(1) values obtained by these 3 methods. In the simulation study, for graphical analysis, the differences between the true and mean of K(1) estimates became larger and the coefficient of variation (COV) of K(1) estimates became smaller as the end time of linear regression became later. The COV of graphical analysis was almost equal to that of NLS with the 1-input compartment model. CONCLUSION: The transfer of (11)C-verapamil from plasma to the brain was able to be quantitatively estimated by graphical analysis because this method can provide K(1) from the data of the initial few minutes without considering the effect of the metabolites in plasma.

ATP Binding Cassette Transporter, Subfamily B, Mem↗

PubMed2001

The effect of ignoring item interactions on the estimated discrimination parameters in item response theory.

Most item response theory models assume conditional independence, and it is known that interactions between items affect the estimated item discrimination. In this article, this effect is further investigated from a theoretical perspective and by means of simulation studies. To this end, a parametric model for item interactions is introduced. Next, it is shown that ignoring a positive interaction results in an overestimation of the discrimination parameter in the two-parameter logistic model (2PLM), whereas ignoring a negative interaction leads to an underestimation of the parameter. Furthermore, it is demonstrated that in some cases the item characteristic curves of the 2PLM and of an item involved in an interaction are quite similar, indicating that the 2PLM can provide a good fit to data with interactions.

Humans↗

PubMed1989

Positron emission tomographic measurement of blood-to-brain and blood-to-tumour transport of 82Rb. I: Error analysis and computer simulations.

Unidirectional blood-to-brain and blood-to-tumour transport rate constants (K1) for 82Rb (half-life 76 s) and plasma water volume per unit mass of brain/tumour tissue (Vp) can be estimated in vivo using dynamic positron emission tomography (PET). The accuracy of these estimates depends upon the accuracy of PET measurements of regional brain/tumour radioactivity and scintillation well detector measurements of whole-blood radioactivity, which, in turn, depend upon the time course of arterial blood radioactivity. A two-compartmental model has been employed to derive estimates for K1, k2 (efflux rate constant) and Vp from 82Rb/PET data. Errors in these parameter estimates have been studied (1) qualitatively using sensitivity function analysis and (2) quantitatively using computer simulations. The effect of adding a third irreversible compartment and its unidirectional rate constant, k3, has also been investigated. The advantages and disadvantages of bolus injection vs continuous infusion protocols are discussed. Precision in estimated parameters from actual patient data is compared to that obtained from computer simulations in part II of this paper.

Blood↗

PubMed2004

Evaluations of maximization procedures for estimating linkage parameters under heterogeneity.

Locus heterogeneity is a major problem plaguing the mapping of disease genes responsible for complex genetic traits via linkage analysis. A common feature of several available methods to account for heterogeneity is that they involve maximizing a multidimensional likelihood to obtain maximum likelihood estimates. The high dimensionality of the likelihood surface may be due to multiple heterogeneity (mixing) parameters, linkage parameters, and/or regression coefficients corresponding to multiple covariates. Here, we focus on this nontrivial computational aspect of incorporating heterogeneity by considering several likelihood maximization procedures, including the expectation maximization (EM) algorithm and the stochastic expectation maximization (SEM) algorithm. The wide applicability of these procedures is demonstrated first through a general formulation of accounting for heterogeneity, and then by applying them to two specific formulations. Furthermore, our simulation studies as well as an application to the Genetic Analysis Workshop 12 asthma datasets show that, among other observations, SEM performs better than EM. As an aside, we illustrate a limitation of the popular admixture approach for incorporating heterogeneity, proved elsewhere. We also show how to obtain standard errors (SEs) for EM and SEM estimates, using methods available in the literature. These SEs can then be combined with the corresponding estimates to provide confidence intervals of the parameters.

Algorithms↗

PubMed2005

Use of reference tissue models for quantification of histamine H1 receptors in human brain by using positron emission tomography and [11c]doxepin.

The aim of the present study is to evaluate the validity of the simplified reference tissue model (SRTM) and of Logan graphical analysis with reference tissue (LGAR) for quantification of histamine H1 receptors (H1Rs) by using positron emission tomography (PET) with [11C]doxepin. These model-based analytic methods (SRTM and LGAR) are compared to Logan graphical analysis (LGA) and to the one-tissue model (1TM), using complete datasets obtained from 5 healthy volunteers. Since HIR concentration in the cerebellum can be regarded as negligibly small, the cerebellum was selected as the reference tissue in the present study. The comparison of binding potential (BP) values estimated by LGAR and 1TM showed good agreement; on the other hand, SRTM turned out to be unstable concerning parameter estimation in several regions of the brain. By including the results of noise analysis, LGAR became a reliable method for parameter estimation of [11C]doxepin data in the cortical regions.

Adult↗

PubMed1998

A pharmacokinetic/pharmacodynamic comparison of SAAM II and PC/WinNonlin modeling software.

This paper presents a detailed comparison of the kinetic analysis software packages SAAM II and PCNonlin/WinNonlin, based on benchmark modeling problems reported in "Pharmacokinetic andPharmacodynamic Data Analysis: Concepts and Applications" (Gabrielsson and Weiner, 1994) and seven additional models. For each model, both software packages were presented with identical implementations. Models were initially executed in PCNonlin or WinNonlin and automated comparisons with SAAM II made using Microsoft Test. Models investigated included one- and multicompartment models with nonlinearities, multiple inputs and samples, multiple simultaneous experiments, and linear equations. Maximum number of compartments, data sets, and parameters were 9, 5, and 10, respectively. We compared 88 different models, many of them in different configurations, e.g., different weighting schemes or different parameter limits. The total number of attempted comparisons between SAAM II and PCNonlin was 161, of which 142 executed without problems. Parameter estimates, their precision (standard errors), and model predictions were compared; a difference of 1% or less was considered "agreement". Observed differences, mainly in parameter standard errors, can be accounted for in terms of different optimization algorithms, convergence criteria, and individual capabilities. In general, there was good agreement (<1% difference) between SAAM II and PCNonlin in terms of parameter estimates and model predictions. However, due to differences in the optimization procedure, parameter standard errors showed considerable differences. Additionally, there were differences when multiple data sets were fitted, indicating the importance of different fitting procedures for interpreting multiple kinetic data sets. The full results of the comparison and the model files in SAAM II and PCNonlin/WinNonlin formats are available from the authors.

Administration, Oral↗

PubMed1998

Optimal design of a two-sample test for assessing [125I]iothalamate plasma clearance in peritoneal dialysis.

BACKGROUND: Plasma clearance of a tracer in peritoneal dialysis (PD) can be used to assess treatment adequacy without labour-intensive fluid collections. Accuracy and precision of plasma clearance estimates by the bolus injection technique depend on the estimation accuracy of the area under the concentration curve and the measurement precision of plasma concentrations. The first source of error is due to oversimplified, e.g. monoexponential, descriptions of plasma disappearance curves. The second source of error arises from the propagation of measurement errors to the parameter estimates. METHODS: The theoretical bias of parameter estimates is determined first for a monoexponential approximation of a biexponential disappearance curve and as a function of the first sampling time at which mixing is still incomplete. The precision of plasma clearance estimates, expressed as coefficient of variation, is then described as a function of the experimental variables and of the standard deviation of measurement error. This allows the determination of the optimal two-sample test that yields most precise estimates of plasma clearance. RESULTS: The optimal two-sample schedules for assessing plasma clearance of [125I]iothalamate in PD patients vary between subjects according to individual clearances and distribution volumes. Our results suggest collecting the first sample 120 min, and the second 2-4 days, after the bolus injection. CONCLUSIONS: The proposed two-sample test is suitable to be used in clinical routine for assessment of adequacy of PD treatment but requires a priori estimation of individual tracer kinetics and of laboratory measurement errors. A fixed design with the first sample taken after 120 min and the second sample collected 3 days after the bolus injection should yield the best performance.

Adult↗

PubMed1983

Experimental artifacts and the analysis of ligand binding data: results of a computer simulation.

Use of computerized analysis techniques for ligand binding data have recently become generally available, and are now used quite routinely. When used appropriately, these tools can improve the precision of the estimated parameters for binding affinity, K, and capacity, R. Furthermore, such programs can also calculate the uncertainty of the estimates e.g., as a percent coefficient of variation (%CV). However, because of unmeasured variability in specific activity, tracer purity, counting efficiency, counter background, efficiency of separation, etc., the actual uncertainty in the parameters K and R is usually much larger than stated. In an attempt to examine the effects of such artifacts, we have developed a computer program which simulates data arising from a number of commonly used experimental designs, and then intentionally distorted with each of these artifacts. Finally, the data are converted to B/F and B and plotted in the conventional Scatchard plot. Distortions revealed in this graph are indicative of the effect each artifact has on the parameter estimates. The computer program is generally written to stimulate the binding of 2 or more ligands to one, two or many classes of independent or cooperative specific sites as well as to nonspecific sites. Thus, the program is applicable in a wide variety of situations. Results show that low tracer purity ("bindability") or low filtration efficiency will significantly alter the measured R value. Poorly determined specific radioactivity may significantly alter the measured K value as well. Imprecise measurement of machine background may result in the spacious appearance of positive cooperativity, or of additional high or low affinity classes of binding sites. Finally, under some circumstances, it is possible to detect and correct for the presence of these artifacts.

Computers↗

PubMedDate not supplied

A simple method for quantifying activity and survival of microorganisms involved in bioremediation processes.

We have developed a substrate-induced growth response (SIGR) method for quantifying activity and population dynamics of microorganisms involved in bioremediation processes in soil and bioreactors. The biomass of organisms that can mineralize a given chemical can be estimated based on the concentration of that chemical needed to induce the growth of the standing population. Estimates of population size are obtained by using nonlinear regression techniques to fit a simple model of microbial population dynamics to biodegradation curves. Using this approach we obtain estimates of values for parameters such as initial population size and growth rate of organisms carrying out biodegradative processes. Our approach was validated by comparing model parameter estimates with independent estimates of the same parameters from the same bioremediation systems. Examples studied include pentachlorophenol degraders introduced into soil and 2,4-dinitrophenol degrading organisms in a bioreactor.

Biodegradation, Environmental↗

PubMed2006

Increased serum interleukin-18 concentration is associated with hypoadiponectinemia in obesity, independently of insulin resistance.

OBJECTIVE: Interleukin-18 (IL-18) is a cytokine with proinflammatory and proatherogenic properties, which might be associated with the development of insulin resistance. In contrast, adiponectin, a protein secreted by adipose tissue, might exert insulin-sensitizing and antiatherogenic effects. The aim of the present study was to analyze the association between serum IL-18 and adiponectin in lean and obese subjects, in relation to insulin resistance. DESIGN: Cross-sectional study. SUBJECTS: One hundred and thirty individuals, 62 lean (body mass index (BMI)<25 kg/m(2), 30 men and 32 women) and 68 with overweight or obesity (BMI>25 kg/m(2), 24 men and 44 women), with normal glucose tolerance and without concomitant diseases. MEASUREMENTS: Oral glucose tolerance test, euglycemic hyperinsulinemic clamp, serum concentrations of IL-18, IL-6, soluble tumor necrosis factor-alpha receptors and adiponectin. RESULTS: Obese subjects had lower insulin sensitivity (M value, P=0.00029) and serum adiponectin (P=0.01) and higher levels of serum IL-18 (P=0.00055). Circulating IL-18 was negatively related to adiponectin (r=-0.31, P=0.00027) and insulin sensitivity (r=-0.33, P=0.00012). Subgroup analysis revealed that these associations were present in the obese (adiponectin, r=-0.38, P=0.0014; M, r=-0.29, P=0.016), but not in lean individuals (r=-0.17, P=0.18 and r=-0.20, P=0.12, respectively). Association of IL-18 with adiponectin remained significant after adjustment for other estimated parameters, including insulin sensitivity. Also, relationship between IL-18 and insulin sensitivity was independent of other estimated parameters. CONCLUSION: Serum IL-18 is inversely related to serum adiponectin, independently of insulin resistance. The relationships of IL-18 with adiponectin and insulin sensitivity are influenced by the presence of overweight/obesity.

Adiponectin↗

PubMedDate not supplied

Principles of good practice for decision analytic modeling in health-care evaluation: report of the ISPOR Task Force on Good Research Practices--Modeling Studies.

OBJECTIVES: Mathematical modeling is used widely in economic evaluations of pharmaceuticals and other health-care technologies. Users of models in government and the private sector need to be able to evaluate the quality of models according to scientific criteria of good practice. This report describes the consensus of a task force convened to provide modelers with guidelines for conducting and reporting modeling studies. METHODS: The task force was appointed with the advice and consent of the Board of Directors of ISPOR. Members were experienced developers or users of models, worked in academia and industry, and came from several countries in North America and Europe. The task force met on three occasions, conducted frequent correspondence and exchanges of drafts by electronic mail, and solicited comments on three drafts from a core group of external reviewers and more broadly from the membership of ISPOR. RESULTS: Criteria for assessing the quality of models fell into three areas: model structure, data used as inputs to models, and model validation. Several major themes cut across these areas. Models and their results should be represented as aids to decision making, not as statements of scientific fact; therefore, it is inappropriate to demand that models be validated prospectively before use. However, model assumptions regarding causal structure and parameter estimates should be continually assessed against data, and models should be revised accordingly. Structural assumptions and parameter estimates should be reported clearly and explicitly, and opportunities for users to appreciate the conditional relationship between inputs and outputs should be provided through sensitivity analyses. CONCLUSIONS: Model-based evaluations are a valuable resource for health-care decision makers. It is the responsibility of model developers to conduct modeling studies according to the best practicable standards of quality and to communicate results with adequate disclosure of assumptions and with the caveat that conclusions are conditional upon the assumptions and data on which the model is built.

Benchmarking↗