Concept Architecture
Concept
Theoretically, Gaussian Quadrature is a numerical integration method that approximates definite integrals by evaluating a function at optimally selected points (nodes) and weighting those evaluations appropriately. Unlike equally spaced numerical integration methods, Gaussian Quadrature chooses both nodes and weights to maximise integration accuracy for a given number of function evaluations. In health economics, it is primarily used to improve numerical approximation of continuous-time processes, particularly within Markov models where it provides more accurate correction for events occurring continuously during model cycles.
Mathematically, Gaussian Quadrature approximates a definite integral as a weighted sum of function values evaluated at Gaussian nodes. The nodes correspond to the roots of orthogonal polynomials, while the associated weights minimise integration error. In health economic modelling, Gauss-Legendre quadrature is most commonly applied for within-cycle correction, providing more accurate estimation of accumulated costs and health outcomes than simpler half-cycle correction methods.
In practice, Gaussian Quadrature is implemented within state-transition models when continuous transitions are approximated using discrete model cycles. Analysts specify the quadrature order, with higher-order methods providing increasingly accurate estimates at the expense of additional computation. Modern health economic software and statistical programming languages frequently implement Gaussian Quadrature automatically for within-cycle correction and numerical integration.
Purpose
Used to improve numerical integration and within-cycle correction in health economic models by providing accurate approximation of continuous costs, health outcomes and transition processes using a limited number of function evaluations.
Mathematical Formulae
Primary Formula
The Gauss-Legendre quadrature approximation is:
???? f(x) dx � ????� w?f(x?)
where:
- x? = Gaussian nodes
- w? = quadrature weights
- n = number of quadrature points
Supporting Formulae
For an arbitrary interval [a, b]:
??? f(x) dx � ((b ? a)/2) ????� w? f(((b ? a)/2)x? + (a + b)/2)
Related Mathematical Methods
- Numerical integration
- Gauss-Legendre quadrature
- Within-cycle correction
- Ordinary differential equation solvers
- Numerical approximation
- Markov cohort modelling
Example
A Markov model uses annual cycles to evaluate a chronic disease intervention over 30 years.
Rather than assuming all transitions occur at the beginning or end of each cycle, fourth-order Gauss-Legendre quadrature is applied to estimate costs and QALYs accrued continuously throughout each year. The numerical integration provides more accurate lifetime estimates than first-order half-cycle correction while requiring relatively few additional function evaluations.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| SUMPRODUCT | =SUMPRODUCT(B2:B5,C2:C5) | Calculate weighted sums of quadrature nodes and weights. |
| INDEX | =INDEX(NodeTable,A2) | Retrieve Gaussian quadrature nodes. |
| INDEX | =INDEX(WeightTable,A2) | Retrieve quadrature weights. |
| POWER | =POWER(A2,2) | Evaluate polynomial or nonlinear functions at quadrature nodes. |
VBA (Optional)
Automate Gaussian Quadrature by generating nodes, applying quadrature weights and evaluating continuous-time model outputs across multiple Markov cycles.
Sources
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation.
- Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed.
- Elbasha EH, Chhatwal J. Theoretical Foundations and Practical Applications of Within-Cycle Correction Methods. Medical Decision Making. 2016.
- Davis PJ, Rabinowitz P. Methods of Numerical Integration. 2nd ed.
- Golub GH, Welsch JH. Calculation of Gauss Quadrature Rules. Mathematics of Computation. 1969.
Related Concepts (6)
Library
Publications
1
Decision Modelling for Health Economic Evaluation — Briggs, Claxton & Sculpher, 1st Edition ed., 2006 (Oxford University Press)
Foundational textbook on decision-analytic modelling for economic evaluation, covering decision trees, Markov models, handling parameter and structural uncertainty, probabilistic sensitivity analysis, and value of information. Volume 1 in the Handbooks in Health Economic Evaluation series.
BookView source →
Frequently Asked Questions (6)
What is Gaussian quadrature?
A numerical integration technique approximating a definite integral using a weighted sum of function values at strategically chosen points.
Source: Press et al. 2007
What problem does Gaussian quadrature solve?
Many quantities in modelling are integrals that cannot be worked out with a formula, such as an expected value taken over a continuous distribution. Gaussian quadrature approximates such an integral numerically by evaluating the function at a small number of carefully chosen points and combining the results with matched weights. Choosing the points well lets it reach high accuracy with few evaluations, which matters when each evaluation is costly. It is a tool for computing integrals that resist exact solution. Press and colleagues (2007) describe the technique.
Source: Press et al. 2007
How does Gaussian quadrature work?
Gaussian quadrature works by approximating an integral as a sum of the function's values at specific nodes, each multiplied by a weight, where the nodes and weights are chosen so the rule is exact for polynomials up to a certain degree. Because both the positions and the weights are optimised, a Gaussian rule with a given number of points achieves higher accuracy than methods using fixed, evenly spaced points. The function need only be evaluated at the chosen nodes to obtain the approximation.
Source: Press et al. 2007
Why is Gaussian quadrature efficient?
Gaussian quadrature is efficient because it chooses the evaluation points and weights optimally, so that a rule with a given number of points integrates polynomials of as high a degree as possible exactly, and approximates smooth functions accurately with few evaluations. Methods that fix the points, such as evenly spaced rules, need more points for the same accuracy. This efficiency matters when each function evaluation is costly, so achieving accuracy with few points saves computation, which is why Gaussian quadrature is widely used.
Source: Press et al. 2007
Where is Gaussian quadrature used in modelling?
In modelling, Gaussian quadrature is used to compute integrals that arise where analytical solutions are unavailable, such as integrating over a distribution of parameters or patient characteristics, computing expected values, or evaluating likelihoods in statistical estimation. In health economic and statistical models, such integrals occur in handling uncertainty and heterogeneity, and Gaussian quadrature provides an accurate and efficient way to approximate them. It is one of several numerical integration methods chosen for their accuracy and efficiency in such calculations.
Source: Press et al. 2007
What are the limitations of Gaussian quadrature?
Gaussian quadrature is most accurate for smooth functions, and it can perform poorly for functions with sharp features, discontinuities, or singularities, where the chosen nodes may miss important behaviour. It also becomes difficult in high dimensions, since the number of points needed grows rapidly with the number of variables, a form of the curse of dimensionality. For such cases, other methods, such as adaptive quadrature or simulation-based integration, may be preferred, so the method is chosen to suit the integral.
Source: Press et al. 2007
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 30 Sep 2025
Content version: 1.0.0
Canonical Identity
- Persistent URI
- https://healtheconomics.wiki/concept/gaussian-quadrature
- Term code
- HE-EM-DM-037
Stable URI · Machine-readable · Resolvable · CC BY 4.0