Intended for healthcare professionals


Commentary: the Dutch approach to unwarranted medical practice variation

BMJ 2011; 342 doi: (Published 17 March 2011) Cite this as: BMJ 2011;342:d1429
  1. Gert P Westert, full professor of health services research,
  2. Marjan Faber, senior researcher
  1. 1Radboud University Nijmegen Medical Centre, PO Box 9101, 114 IQ Healthcare, Nijmegen, 6500 HB, Netherlands
  1. Correspondence to: G Westert g.westert{at}

Medical practice varies considerably, both within and between countries. Rates of surgical procedures and medical treatment seem to be unrelated to illness and other patient related factors in many studies.1 Wennberg makes the intriguing observation that supply is the prime determinant of healthcare usage in the United States: most unwarranted healthcare is given in areas with high supply levels and does not result in improved patient outcomes. 1

Wennberg defines three categories of care: effective care, preference sensitive care, and supply sensitive care, with supply factors relating to overuse in the last two categories. Supply levels in the Netherlands vary less than in the US. This is because healthcare supply in the Netherlands was—at least until 2006—centrally planned and controlled by the government. You might therefore expect that variations in delivery of healthcare would be insignificant. Unfortunately, this is not the case. The 2010 Dutch healthcare performance report notes remarkable variations in quality and price of healthcare between care providers.2 For instance, the percentage of unplanned caesarean sections carried out in low risk pregnancies varied from 7.3% to 30.2% across hospitals. This variation is unwarranted because it is to a large extent unrelated to patient based factors.

In preference sensitive care there is scientific uncertainty about the best option for effective treatment in terms of mortality or quality of life; at least two equipoise treatment options exist.1 In these cases the decision is theoretically shared between the patient and doctor. However, most patients still leave the decision to the doctor. As a consequence, the supplier dominates.

Shared decision making in preference sensitive care is still a novelty in the international healthcare sector. Recent Consumer Quality Index questionnaire data on experiences of Dutch hospital patients shed some light on this.3 4 Only 50% of patients with a malignant breast condition (n=393) or spinal disc herniation (n=1521) reported that they were fully involved in decision making about treatment and care in a survey conducted by the independent non-profit organisation, Consumer Experience Centre (CKZ).4 A fifth of patients said they were “never/ sometimes” involved (fig 1). An enormous challenge here is to enhance the patient’s role in determining the use of preference sensitive care. Variation in these cases is unwarranted if it is predominantly doctor driven and not related to patients’ preferences.


Percentage of care users who reported that they were involved in decisions about care and treatment, 2005-84

Since the introduction of managed competition into the Dutch health system in 2006,5 the volume and fees of elective surgery are set by free negotiation between health insurers and providers. As might be expected, given the fact that shared decision making is still a novelty, the volume of preference sensitive care went up rapidly and regional variation increased. The number of cataract procedures, for example, increased by roughly 25% between 2005 and 2010.6 Furthermore, in areas with centres that specialised in certain procedures, the rates of those procedures were much higher than in other areas.7 Apparently the indication for a surgical procedure varies substantially across hospitals. Since the Dutch prefer to get their care nearby, geography seems to matter: what you get depends on where you live. Similar results were observed for prostatectomies and tonsillectomies.

The final category Wennberg mentioned is supply sensitive care—everyday care used by people with acute and chronic conditions (physician visits, referrals, prescription of drugs, tests, etc). In general the more this is supplied and easy accessible, the more that is consumed. Wennberg states that more of this care is in most cases not related to better outcomes. Primary care providers largely determine the frequency of such care. In the Netherlands, where general practitioners act as the gatekeeper of the healthcare system, bundled payment was recently introduced for chronic diseases—for example, diabetes, chronic obstructive pulmonary disease, and management of cardiovascular risk.8 Under this system insurers pay a single fee to a newly created contracting entity—the “care group”—to cover a full range of care for a fixed period. The care provided is defined by national guidelines. General practitioners have taken a central position in the care groups. The new system is to bring down unwarranted variation and fragmentation of care. Early signs show that the delivery of diabetes care has improved as a result of the enhanced coordination of care, but it is too soon to see differences in outcomes (such as avoidable hospital admissions).


Cite this as: BMJ 2011;342:d1429


  • doi:10.1136/bmj.d1513
  • Competing interests: All authors have completed the unified competing interest form at (available on request from the corresponding author) and declare no support from any organisation for the submitted work; no financial relationships with any organisation that might have an interest in the submitted work in the previous three years; and no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract