Dr. Muijen is regional adviser for mental health for the World Health Organization's Regional Office for Europe, Scherfigsvej 8 DK-2100, Copenhagen, Denmark (e-mail: mfm@euro.who.int).

The brief report by Priebe and colleagues (1) attempts to demonstrate that institutional care for people with severe mental disorders has increased in several Western European countries during the 21st century, after several decades of deinstitutionalization in the 20th century. The discussion guardedly suggests that such a trend is occurring, implying an underlying political shift toward more restrictive and risk-averse forms of care.

Some caution is advised. First, the term "institutionalization" carries many meanings. The lack of a specific definition in the report by Priebe and colleagues matters. Nobody would deny that people need houses to live in and beds to sleep in. It is unhelpful to group all places available for people with mental disorders and categorize these as institutions. Data should differentiate between "community-based facilities" that offer personalized interventions, which likely would include residential facilities, and "institutional places" that impose standardized care.

Central to the analyses of the numbers in this report are the time points. The baseline is 1990, which arguably was the end of the Cold War. However, 1990 is not significant as a time point for bed closures. The number of psychiatric beds peaked in the 1950s in most countries, with a rapid fall in the subsequent decades. The challenge of the 1990s, in Western Europe as well as in the United States, was that the decline in the number of hospital beds was not compensated for by the development of community services, including residential care facilities, which left patients who would previously have been institutionalized at high risk of destitution. Since then many countries have invested large sums to support the development of nursing homes and residential care facilities; such investment has occurred at a fast pace, as this brief report illustrates. This line of reasoning points to another conclusion—namely, that we are not dealing with an overdevelopment of beds but, as a result of inadequate planning, with a lag in time between hospital closures and the availability of residential places.

The increase in the number of forensic beds in some countries is a different issue. Such beds were established by governments in response to public pressure resulting from high-profile incidents and an inadequate supply of beds after the closure of psychiatric hospitals. The overall figures hide a shift in several countries from a small number of large prison hospitals toward smaller-scale specialty psychiatric units. Also, it should be recognized that even though the number of forensic beds has increased substantially in a few countries, these places make a relatively small contribution to the overall number of psychiatric beds.

Relevant to the argument is the increase in the number of prison beds. Although we know that a sizeable number of prison inmates have mental illnesses, mostly personality disorders, the association between hospital closures and the presence of people with mental illness in prisons is very complex. It would be highly speculative to estimate the number of persons now in prison who would in the past have lived in mental institutions.

The strongest point made by Priebe and colleagues is in regard to the diversity across even fairly homogeneous Western European countries—no single variable is moving in a consistent direction. All countries are developing complex ranges of services, attempting to shift from supply-driven hospital care to needs-led community-based services that are supported by a range of specialist "institutions." These authors rightly acknowledge the absence of any convincing trends apart from an increase in residential care.

Of course, we need to guard against slipping back to widespread institutionalization, and this report serves an important purpose in keeping us aware. However, a more urgent challenge is alluded to at the end of the report. By chance, just before reading the report, I read brief accounts of two incidents in the March 29, 2008, edition of The Daily Telegraph, a British newspaper. The first involved the manager of a care home who was fired because she neglected her residents. Immediately above this article was the story of a husband who let his wife starve to death while they were living together. She was unable to care for herself because of a mental illness. These are the unintentional consequences of decentralized mental health care and higher levels of patient autonomy. Mental health care has moved beyond care in beds. Mental health service providers are now facing the urgent challenge of how to formulate effective strategies to prevent both institutional dependence and community neglect and of how to close the gap between these two poles.