Advisory cycle / Letter on utilisation of long-term care budget framework

Dutch Healthcare Authority (NZa)

To forecast spending under the Long-Term Care Act (LTC), an algorithm is deployed. Based on aggregated data, that rhythm shows whether the available budget framework is sufficient.

Last change on 10th of December 2025, at 13:56 (CET) | Publication Standard 1.0

• Publication category: Other algorithms
• Impact assessment: Privacy Quickscan
• Status: In use

General information

Theme

Health and Healthcare

Begin date

05-2018

Contact information

https://www.nza.nl/contact

Link to publication website

https://puc.overheid.nl/nza/doc/PUC_787334_22/

Responsible use

Goal and impact

The purpose of the letters is to forecast expenditure under the Wlz for the year in question (year t). The result is a letter in which we explain this forecast and show the extent to which we expect the budget framework to be adequate. An important element of the letter is the report in which we show the comprehensive results of the various scenarios at Wlz administrator level.

The impact is that partly on the basis of the letter sent to the minister, the minister decides on the use of redistribution funds. In addition, the minister may decide partly on the basis of our projections to possibly add other additional resources to (sub-)frameworks.

Considerations

The Ministry of Health, Welfare and Sport annually sets the budget framework for the Long-Term Care Act (Wlz). In June of year t-1, VWS indicates the provisional framework for the following year in the Preliminary Framework Letter. The starting point of the framework is always the previous year's framework. This framework is adjusted based on a growth allowance for demographic developments and increasing care severity, expected wage and price developments, and policy changes. In September/October year t-1 follows the Final Wlz Framework Letter in which VWS indicates what the final framework looks like.

The budget framework consists of:

- The contracting space for care in kind (zin);

- The available amount for personal budgets (pgb);

- The reallocation resources: resources that can be used to absorb unexpected shifts in care provided. These resources are withheld by VWS at the start of the year and only deployed when necessary.

At the request of the Minister of VWS, the NZa monitors the utilisation of the Wlz framework. The minister can then decide to deploy the redistribution resources or other additional resources.

Human intervention

The minister can disregard the contents of the letter on the adequacy of the framework and decide independently whether or not to raise the framework. This is a political consideration.

The projections initially made by the NZa are consulted with healthcare offices. Adjustments may follow through 'expert opinions' if outcomes do not seem plausible.

Risk management

The risks in terms of technology, legislation, costs, ethics, explainability and other issues are very limited. The algorithm is not critical, no direct decisions follow from it and the technique is understandable and explainable.

Links to legal bases

Art. 49e Wmg: https://wetten.overheid.nl/jci1.3:c:BWBR0020078&hoofdstuk=4&paragraaf=4.4&artikel=49e&z=2025-01-01&g=2025-01-01

Elaboration on impact assessments

A legitimacy assessment has been done on the processes in which this algorithm plays a role according to our standard procedure. The impact on privacy of this algorithm is negligible.

Impact assessment

Privacy Quickscan

Operations

Data

The following data sources are used for the advice:

Indications CIZ: Monthly delivery of indications CIZ

Care-in-kind

1. Monthly delivery Vektis Care-in-natural

Personal budget (PGB)

2. Personal budget directly from Vektis monthly, pseudonym not linkable

Waiting list data Care institution Netherlands (ZiNL)

Technical design

The advisory cycle uses data aggregated by care office on a monthly performance basis. This data is further aggregated at Wlz-employer level, from which a time series is produced per Wlz-employer per performance on a monthly basis. An ets model is applied to this time series, predicting the demand for the relevant performance over the months in the year for which an advice is requested.

Tracking the development in PGB costs is done by aggregating the data on a monthly basis per care office and creating a time series from this. An ets model is fitted to this time series, predicting the expected budget required during the year for which advice is sought.

As a second scenario, the advice cycle looks at the growth of indications per care profile over the past year. This growth is extended linearly to predict the growth of the year for which advice is sought. A price tag is attached to the growth of an indication per care profile using the aggregated data obtained from the distribution model. The split of an indication into cashing in ZIN/PGB is done proportionally as the split ZIN/PGB is currently known.

The waiting list data are summarised to provide insight per care office and sector into the waiting status and duration. If necessary, a financial value can be attached to this at the request of the Minister of LZ. A value is then assigned per care profile to the waiting lists from the aggregated data obtained from the distribution model.

External provider

Scenario obv declarations: ETS algorithm https://otexts.com/fpp3/ets-forecasting.html. For model fit: function 'ets' from forecast package of R https://pkg.robjhyndman.com/forecast/