Admin time in Dutch healthcare, the AZWA target and how it is measured

For the Labour Market in Health and Social Care research programme, Statistics Netherlands (CBS) uses an employee survey to measure how much working time goes on reporting on the care delivered and on record-keeping for the organisation. For the second quarter of 2025, published on 18 November 2025, the two together came to almost a third of working time. For client-facing occupations it is around 38 per cent in youth care, around 36 per cent among GPs and in hospitals, and around 28 per cent in nursing, residential and home care. Since 2021 the total has hovered around 30 to 31 per cent. More than 44 per cent of client-facing employees experience high or very high work pressure; among those who spend an above-average amount of time on admin, it is 50 per cent.

Since 2016 Berenschot has asked professionals in long-term care the same question, voluntarily and anonymously, and has collected more than 18,500 completed surveys since then. In the 2025 round, 1,952 people took part: 1,042 from nursing, residential and home care, 745 from disability care and 165 from mental healthcare. They spend an average of 34 per cent of their time on admin and consider 23 per cent acceptable. Eight in ten find these tasks a burden; the electronic record, reports and meetings take up the most time. More than half say their administrative workload increased in the past year, while the measured time is falling slightly. Berenschot finds this hard to explain and wants to look into it further. Those who feel they have more say over their own work accept a larger share of admin.

The AZWA, the Dutch supplementary agreement on health and social care, sets the target at no more than 20 per cent admin time by 2030. In answers to the Dutch House of Representatives of 9 June 2026, the ministers cite the latest CBS figure: 30.1 per cent at the end of 2025, against 31.0 per cent at the end of 2019. No annual targets have been agreed. The government expects the decline to speed up later, partly because sector and professional associations put forward ten burdensome requirements and authorisations each year, which are then tackled. On 24 June 2026, MP Coenradie tabled a motion asking for a uniform measurement method, a baseline measurement and, for each sector, a timetable with measures, responsible parties and reporting points before the end of 2026.

High hopes are pinned on speech technology in the consulting room, writes M&I/Partners in the AI Monitor Ziekenhuizen of March 2026, based on responses from 52 hospitals. Of these hospitals, 69 per cent have started with voice-driven reporting. Most are at the pilot stage, 10 per cent are implementing and 17 per cent are scaling up. On average, 21 per cent of the healthcare providers it is intended for actively use the technology. The hospitals describe the impact AI has actually had over the past three years as neutral.

How solid is the evidence? Two independent series, CBS since 2019 and Berenschot since 2016, have shown the same level for years, and the government confirms the figure in parliament. The gap with the target is a given to plan around. Whether it will still close in the years to 2030 is open, and without a fixed measurement method that will also be hard to establish.