Concept Architecture
Concept
Theoretically, Parametric Estimation is a statistical estimation approach in which the probability distribution of the data is assumed to belong to a specified family characterised by a finite set of parameters. It is founded on parametric statistical theory and enables inference by estimating distributional parameters such as means, variances, regression coefficients or shape parameters. The approach exists to provide efficient statistical estimation when the assumed probability model adequately represents the underlying data-generating process.
Mathematically, parametric estimation involves specifying a probability density or probability mass function indexed by unknown parameters and estimating those parameters from observed data. Maximum likelihood estimation is the most widely used estimation framework, although least squares and Bayesian estimation are also common depending on the modelling context. Estimated parameters define the fitted probability distribution or statistical model.
In practice, parametric estimation is applied across regression analysis, survival analysis, generalised linear models and decision modelling. Model parameters are estimated using statistical software, with model adequacy assessed through likelihood-based statistics, residual diagnostics and information criteria. In health economics, parametric estimation underpins the estimation of survival models, cost models, utility models and disease progression models used in economic evaluation.
Purpose
Used to estimate the parameters of predefined statistical models, enabling prediction, inference and extrapolation in health economic analyses.
Mathematical Formulae
Primary Formula
?? = arg max L(?)
where:
?? = estimated parameter vector
L(?) = likelihood function
Supporting Formulae
L(?) = ????� f(x? | ?)
?(?) = ln(L(?))
?? = arg max ?(?)
where:
? = parameter vector
f(x? | ?) = assumed probability distribution
?(?) = log-likelihood function
Related Mathematical Methods
- Maximum Likelihood Estimation
- Least Squares Estimation
- Bayesian Estimation
- Parametric Survival Analysis
- Generalised Linear Models
- Regression Analysis
Example
A Weibull survival model is fitted to oncology trial data. The model assumes survival follows a Weibull distribution and estimates the shape parameter as 1.42 and the scale parameter as 18.7 months using maximum likelihood estimation. These estimated parameters are subsequently used to extrapolate long-term survival within a cost-effectiveness model.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| LN | =LN(B2) | Calculate log-likelihood components. |
| SUM | =SUM(C2:C201) | Calculate the total log-likelihood. |
| Solver | Maximum likelihood optimisation | Estimate model parameters by maximising the likelihood function. |
| EXP | =EXP(D2) | Transform estimated parameters where required by the model. |
VBA (Optional)
Automate maximum likelihood estimation for alternative statistical models and compare competing parameter estimates using likelihood-based criteria.
Sources
- Casella G, Berger RL. Statistical Inference.
- Pawitan Y. In All Likelihood: Statistical Modelling and Inference Using Likelihood.
- Collett D. Modelling Survival Data in Medical Research.
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation.
- NICE. Health Technology Evaluation Manual.
- McCullagh P, Nelder JA. Generalized Linear Models.
Related Concepts (2)
Library
Publications
1
NICE DSU Technical Support Document 14: Survival analysis for economic evaluations alongside clinical trials – extrapolation with patient-level data — Nicholas R. Latimer, TSD 14 ed., 2013 (NICE Decision Support Unit (University of Sheffield))
The reference guidance on survival analysis for economic evaluation: fitting standard parametric models (exponential, Weibull, Gompertz, log-logistic, log-normal) to censored trial data and extrapolating to estimate lifetime survival benefit, with a process guide for model selection and justification.
Frequently Asked Questions (6)
What is parametric estimation?
A statistical approach estimating a survival function by assuming the data follow a specific distribution, such as Weibull or log-normal, with estimated parameters.
Source: Collett 2015
What does parametric estimation trade for the ability to extrapolate?
By assuming the data follow a chosen distribution, parametric estimation reduces the whole survival curve to a few parameters and gains a formula that can be extended beyond the observed period. The price is a risk that the assumed distribution does not match reality, so a poorly chosen form can fit the data adequately yet extrapolate badly. It trades the fidelity of making no assumption for the power to project and to summarise with few numbers. The gamble rests on choosing an appropriate distribution. Collett (2015) describes this trade-off.
Source: Collett 2015
How does parametric estimation work?
Parametric estimation works by choosing a survival distribution, then estimating its parameters from the data, usually by maximum likelihood, which finds the parameter values making the observed data most probable under the distribution, accounting for censoring. The fitted distribution gives a smooth survival and hazard function defined by the estimated parameters. Because the distribution has a known form, the fitted curve can be evaluated at any time, including beyond the observed follow-up, enabling extrapolation from the estimated parametric model.
Source: Collett 2015
What are the advantages of parametric estimation?
Parametric estimation gives a smooth survival curve summarised by a few parameters, allows the hazard and survival to be evaluated at any time, and, crucially, permits extrapolation beyond the observed data, which non-parametric methods cannot do. This makes it necessary for estimating long-term and mean survival for economic evaluation. It can also be efficient when the assumed distribution is correct. These advantages make parametric estimation central to survival modelling where projection beyond the data is required.
Source: Latimer 2013
What are the limitations of parametric estimation?
Parametric estimation depends on the assumed distribution being appropriate, so an incorrect choice biases the fit and, especially, the extrapolation, since the projected survival follows the assumed form beyond the data where it cannot be checked. Different distributions fitting the observed data similarly can extrapolate very differently, making long-term estimates uncertain. The assumption of a particular form is a strong one. These limitations mean the distribution is chosen carefully, its fit and extrapolation assessed, and sensitivity to the choice examined.
Source: Latimer 2013
How does parametric estimation differ from non-parametric estimation?
Parametric estimation assumes a specific distribution and fits its parameters, giving a smooth curve that can be extrapolated, whereas non-parametric estimation makes no distributional assumption, estimating survival directly from the data as a step function that cannot be extended beyond it. Parametric estimation enables projection at the cost of a distributional assumption, while non-parametric estimation is robust and honest within the data but limited to it. The two are complementary, with non-parametric methods describing observed survival and parametric methods extrapolating.
Source: Collett 2015
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 22 Oct 2025
Content version: 1.0.0
Canonical Identity
- Term code
- HE-EM-SM-060
Stable URI · Machine-readable · Resolvable · CC BY 4.0