Concept Architecture
Concept
Theoretically, Piecewise Model is a mathematical modelling approach in which different mathematical functions are defined over distinct regions of the domain, with each function representing behaviour within a specified interval or condition. Piecewise modelling exists because many biological, clinical and economic processes exhibit structural changes that cannot be adequately represented by a single continuous function. In health economics, piecewise models are commonly applied to survival analysis, hazard functions, cost trajectories and disease progression where different phases require different mathematical representations.
Mathematically, a piecewise model is represented as a function composed of multiple sub-functions, each applying within a defined interval. The mathematical framework partitions the domain into mutually exclusive regions and specifies the corresponding functional form for each region. Parameters may differ between segments while continuity may or may not be imposed at transition points depending on the modelling assumptions.
In practice, piecewise models are estimated by identifying clinically or statistically justified breakpoints and fitting separate models within each interval. In health economic evaluation they are frequently implemented as piecewise exponential survival models, piecewise linear cost functions or segmented treatment-effect models to improve model fit where hazards or costs change over time.
Purpose
Used to represent processes whose mathematical behaviour changes across predefined intervals, improving the accuracy of survival models, economic evaluations and disease progression models when a single functional form is inadequate.
Mathematical Formulae
Primary Formula
f(x) = { f?(x), x < c? f?(x), c? � x < c? ? f?(x), x � c??? }
Supporting Formulae
Piecewise linear model:
y = { ?? + ??x, x � c (?? + ??c) + ??(x ? c), x > c }
Piecewise exponential hazard:
h(t) = ??, �??? � t < �?
Survival function:
S(t) = exp(???(t ? �???))
Related Mathematical Methods
- Piecewise Exponential Model
- Segmented Regression
- Change-Point Analysis
- Survival Analysis
- Maximum Likelihood Estimation
- Spline Regression
Example
A health economic evaluation models overall survival after cancer treatment using a piecewise exponential model. The estimated hazard is 0.03 per month during the first 12 months and 0.08 per month thereafter.
For month 8:
h(t) = 0.03
For month 18:
h(t) = 0.08
The piecewise model captures the increase in mortality risk after the first year and provides a better fit than a model assuming a constant hazard throughout follow-up.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| IF | =IF(A2<=12,0.03,0.08) | Apply interval-specific hazard rates |
| IFS | =IFS(A2<=6,Rate1,A2<=12,Rate2,TRUE,Rate3) | Implement multiple piecewise intervals |
| INDEX | =INDEX($B$2:$B$5,MATCH(A2,$A$2:$A$5,1)) | Retrieve parameters for each interval |
| MATCH | =MATCH(A2,$A$2:$A$5,1) | Identify the appropriate interval |
| EXP | =EXP(-B2*C2) | Calculate interval-specific survival probabilities |
VBA (Optional)
A VBA procedure can automatically allocate observations to predefined intervals and apply the appropriate model parameters throughout a health economic simulation.
Sources
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation.
- Collett D. Modelling Survival Data in Medical Research.
- ISPOR-SMDM Modeling Good Research Practices Task Force Reports.
- NICE. Health Technology Evaluation Manual.
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 a piecewise model?
A survival model dividing follow-up into distinct segments, each modelled with a separate, often simpler function, letting the hazard shape vary in steps.
Source: Collett 2015
What can a piecewise model capture that a single distribution cannot?
A single distribution imposes one hazard shape across the whole of follow-up, but risk sometimes behaves quite differently in different periods, high just after treatment, low for a stable middle stretch, then rising again. A piecewise model captures this by dividing time into segments and fitting a separate, often simple, hazard within each, so the overall shape can change abruptly at the boundaries. This lets it follow patterns that no single smooth form matches. The segments are chosen where the hazard is thought to change. Collett (2015) describes this approach.
Source: Collett 2015
How does a piecewise model work?
A piecewise model works by splitting the time axis into segments at chosen cut points and fitting a separate hazard function, often a constant or simple form, within each segment. The hazard can jump between segments, so the overall hazard follows a step or piecewise pattern that can approximate a changing hazard. Parameters are estimated for each segment from the data falling in it. By combining the segment-specific functions, the model represents a hazard that varies over time without assuming one smooth distribution.
Source: Collett 2015
When is a piecewise model used?
A piecewise model is used when the hazard changes over time in ways a single parametric distribution captures poorly, and a flexible, segmented representation is wanted, or when different phases of follow-up have distinct risk patterns, such as high early risk after surgery then a lower stable phase. It is also used in extrapolation to model the observed period and a separate long-term hazard. The piecewise approach suits situations where dividing follow-up into intervals with separate hazards better represents the changing risk.
Source: Latimer 2013
What are the advantages of a piecewise model?
A piecewise model is flexible, able to approximate complex or changing hazards by combining simpler functions over segments, without committing to a single parametric form that may fit poorly. It is relatively easy to understand and fit, since each segment uses a simple function, and it can represent distinct phases of risk. In extrapolation, it allows the long-term hazard to be specified separately from the observed period. This flexibility makes piecewise models useful where the hazard varies in ways standard distributions cannot capture.
Source: Collett 2015
What are the limitations of a piecewise model?
A piecewise model requires choosing the number and location of the cut points dividing follow-up, which affect the fit and can be somewhat arbitrary, and the hazard jumps between segments, giving a discontinuous rather than smooth pattern. Too many segments risk overfitting, and the choice of the final segment's hazard strongly affects extrapolation. As with other models, its projection beyond the data is uncertain. These limitations mean the cut points and segment hazards are chosen carefully and their influence on results examined.
Source: Latimer 2013
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 23 Oct 2025
Content version: 1.0.0
Canonical Identity
- Persistent URI
- https://healtheconomics.wiki/concept/piecewise-model
- Term code
- HE-EM-SM-066
Stable URI · Machine-readable · Resolvable · CC BY 4.0