To avoid misinterpretation of results, research based on administrative data requires extensive knowledge of the context and background of information on service use. In a national study as part of the periodic evaluation of the Act on Special Admissions to Psychiatric Hospitals, Scholten and Tjadens (1996) found that in about 25% of the cases a judge would not grant authorization to continue the compulsory admission that in emergency situations is issued by the city mayor.

The authors reported striking regional differences and the Rotterdam area beat all others: 42% of emergency compulsory admissions were not continued. Scholten and Tjadens, however, did not account for differrences in the administration of compulsory admissions that were not continued. They probably included patients who were hospitalized outside the residence area, as a consequence of which the admission procedure is discontinued in that area and the judiciary in another area takes over. At the time, relatively many Rotterdam patients were admitted elsewhere because of a shortage of hospital beds in the Rotterdam-Rijnmond district. In a replication study we found no more than 20% discontinued involuntary admissions in the city of Rotterdam (Wierdsma 2003).

Scholten, CM, and FLJ Tjadens. 1996. Wet BOPZ: Verkommerden en verloederden. [Dutch Act on Special Admissions to Psychiatric Hospitals: The Withered and Degenerated] Rijswijk. Ministerie van Volksgezondheid, Welzijn en Sport.

Wierdsma, AI. 2003. "Inbewaringstelling... en dan? Een Rotterdams onderzoek naar de motieven van Justitie om gedwongen opnemingen niet voort te zetten." [A detention order... and what next? A Rotterdam study on why prosecutors and courts decide against compulsory admission.] Maandblad Geestelijke volksgezondheid 58: 337-349.

Register research requires detailed insight into administrative processes and procedures in mental healthcare. In a study of the effects of the implementation in 1994 of the new Act on Special Admissions to Psychiatric Hospitals on the development of the number of emergency compulsory admissions, Poletiek (1997) over-counted patients admitted involuntarily. In the period 1990–1995 a steady increase in the number of involuntary admissions was found. However, particularly after the introduction of the new law, there was a noticeable increase in the number of emergency compulsory admissions of patients who were already admitted voluntarily. The increase of these ‘conversions’ was analyzed based on the bed-occupancy figure per year, broken down by judicial status.

This occupancy figure is an estimate of the number of patients on any day of the year and whose stay in the hospital is involuntary. That number was overestimated because judicial status was often not updated following the expiration of the emergency compulsory procedure. In this way, the bed-occupancy figure included patients who were treated voluntarily after compulsory admission of three weeks maximum. The increase of ‘conversions’ after the implementation of the new Special Admissions Act could be interpreted as an effect of a stronger awareness of the Registration of judicial measures (Wierdsma 1997). This does not mean that the judicial status would be more frequently updated at a later time, as Poletiek’s reply to the above criticism suggested. Instead, stronger registration awareness could account for more changes in judicial status, all of which would be included in the occupancy figure. Bias in the national register data was confirmed, yet Poletiek held to the conclusion that the number of involuntary admission in psychiatric hospitals increased after the introduction of the new law.

Poletiek, FH. 1997. "De Wet BOPZ getoetst aan de cijfers." [The formal Admissions to Psychiatric Hospitals Act (BOPZ) tested on the basis of published figures.] Maandblad Geestelijke volksgezondheid 4: 349-361.

Wierdsma. AI. 1997. "BOPZ cijfers getoetst " [Evaluating indicators of the Act on Special Admissions in Psychiatric Hospitals.] Maandblad Geestelijke volksgezondheid 52: 1024-26.

In some cases the epidemiological numerator or denominator is miscalculated, which leads to statistical bias. In his dissertation study, Blansjaar (1992) calculated high rates of Korsakoff patients in The Hague: 214 patients were registered in various mental healthcare services, so that relative to the population of The Hague the prevalence-rate was estimated at almost 4,8 per 10.000 inhabitants.

Based on admission data in the Province of North-Holland, Schnabel (1992) estimated the Korsakoff syndrome prevalence-rate considerably lower: about 3 patients per 10,000 inhabitants. Prof. Paul Schnabel later became director of the Netherlands Institute for Social Research, so ... (argumentum ad hominem).

Blansjaar’s estimates were probably biased, because it is unclear if before admission all patients in his study in The Hague had lived in the city. The Hague had a specialized sheltered home for Korsakoff-patients, which most likely had a regional instead of a local function. In that case, not the urban population but inhabitants from a wider area should have been the epidemiological denominator, which means that prevalence of the Korsakoff-syndrome was overestimated.

Blansjaar, BA. 1992. Alcoholic Korsakoff's Syndrome.Thesis Leiden. Rijksuniversiteit Leiden