In general, psychiatric treatment in Scandinavian countries, including Denmark, is public and open to all residents. All expenses are paid through taxes. Over the past ten to 15 years, private hospitals have been established in Denmark, although not yet for adult psychiatry. Approximately 15% of psychiatrists in Denmark are in private practice, and virtually all of them have an agreement with the public health insurance system regarding reimbursement and the services that may be rendered by psychiatrists in private practice. In the past few years, some psychiatrists have established private practices without having an agreement with the public health insurance system, and they are paid by the patients or by private insurance companies.
Most psychiatrists who are in private practice in Denmark work exclusively in private practice, whereas most U.S. psychiatrists in private practice also work in hospitals (1). The Danish hospital-based, public psychiatric system has the responsibility for the most severely ill patients who need inpatient treatment, long-term follow-up, home visits, and integrated social psychiatric care. Psychiatrists in private practice offer exclusively office-based outpatient treatment, which is reflected in the diagnoses of patients they treat. Danish studies from before decentralization —when the ICD-8 was still used—found that approximately 60% of patients treated by psychiatrists in private practice had a diagnosis of neurosis (2). A Danish pilot study from 2006 found that a very small proportion—3.4%—of patients treated in private practice had a psychotic disorder, compared with 82.6% with affective, anxiety, and personality disorders (3). The proportion of patients treated in private practice who have affective disorders has been found to range from 25% to 50% in the United States (4,5,6), Canada (7,8), and France (9). These studies found a range of 10% to 33% for anxiety disorders and 10% to 40% for personality disorders.
Currently, about 150 psychiatrists in Denmark are in private practice, compared with 100 in the mid-1990s. It is not known whether this increase and the decentralization of hospital-based psychiatry, which began in 1980, have changed the diagnostic profile of patients treated in private practice to include a larger proportion with psychotic disorders. In addition, studies in Denmark in the 1980s found that rates of treatment dropout among patients treated by psychiatrists in private practice differed by diagnosis: from 18% to 33% of patients with personality disorders, 16% of patients with neuroses, and 8% of patients with psychoses (10,11). A 2006 pilot study found an overall treatment dropout rate in private psychiatric practice of 24.4% (3).
The aim of this study was to investigate the diagnostic profile of patients treated in Danish private psychiatric practice and to identify predictors of treatment dropout in this setting.
In 1996 a Danish group of psychiatrists in private practice established a clinical database—Quality Assurance in Danish Private Psychiatric Practice—to accumulate data for continuous quality assurance and research in this setting (12). Psychiatrists in private practice who have an agreement with the public health insurance system submit data. None are child and adolescent psychiatrists. Psychiatrists can participate only if they report data for all treatment episodes for each year of participation.
A treatment episode is defined as the time from the first visit until official discharge. A referral from the patient's general practitioner is required. An episode consists of a number of visits and telephone contacts between the patient and psychiatrist; the number is not established in advance. For a discharge to be official, the patient and the psychiatrist must agree that no more consultations are planned, and a discharge letter is then sent to the patient's general practitioner. In case of dropout, the general practitioner is informed about the dropout and termination of treatment. After terminattion, a new episode can be initiated as soon as the patient is referred again by the general practitioner—no hiatus period after discharge is warranted for the episode to be considered "new."
All episodes are registered, and data include the referral date, date of each contact (in person and telephone), modality (for example, psychotherapy or psychopharmacologicy), date and type of termination (official discharge or dropout), and ICD-10 diagnoses. The study met regulations of the Danish Data Surveillance Agency and was approved by the Ethical Committee of the County of Aarhus.
From 1996 to 2006, a total of 37 psychiatrists reported to the database (mean±SD number per year 18.1± 2.9; range 14–22). As noted above, the number of psychiatrists in private practice increased over this period from approximately 100 to 150; thus approximately 15% of psychiatrists in private practice reported to the database each year. A total of 41,462 episodes were initiated over the period; in 14,861 (35.8%) the patients were men, and in 26,601 (64.2%) they were women. At treatment initiation, the mean age was 47.5±15.1. A total of 35,205 patients accounted for the 41,462 episodes—12,670 (35.9%) men and 22,535 (64.0%) women. For each patient the mean number of episodes was 1.18±.49.
For the 41,462 episodes, affective disorders (ICD-10 F3) accounted for the largest proportion of diagnoses—39.0% of episodes for men (N=5,796) and 41.3% for women (N=10,994). This was followed by nervous and stress-related disorders (ICD-10 F4 group)—29.5% for men (N=4,384) and 34.8% for women (N=9,265)—and then by personality disorders (ICD-10 F6)—10.6% (N=1,575) and 10.0% (N=2,666), respectively. Organic diseases, substance use disorders, and schizophrenia (ICD-10 F0, F1, and F2, respectively) together accounted for 20.9% of the episodes for men (N=3,106) and 13.8% of those for women (N=3,676). Schizophrenia spectrum disorders accounted for 4.1% of episodes for men (N=609) and 2.5% for women (N=665).
Of the 41,462 episodes, 33,494 (80.8%) had been terminated by the end of the study. For these episodes, the mean number of contacts was 7.1±9.3, and for 57.9% of these episodes (N=19,405) there were four or more contacts. The three largest diagnostic groups—affective disorders, nervous and stress-related disorders, and personality disorders—accounted for 78.9% (N=26,443) of the 33,494 terminated episodes, with 7.3± 9.3 contacts per episode and 59.8% of these episodes (N=15,813) with four or more contacts.
Of the 33,494 terminated episodes, 25.9% (N=8,199) ended in dropout. For these episodes, there were 5.7±6.8 contacts per episode, with 49.9% (N=4,091) involving four or more contacts. Of the 26,443 terminations accounted for by the three largest diagnostic groups, 26.2% (N=6,924) ended in dropout, with 5.8±6.7 contacts per episode and 51.4% (N=3,572) involving four or more contacts.
Table 1 presents results from univariate and backward stepwise regression analyses of predictors of dropout. Younger age, male sex, personality disorder (versus affective or nervous and stress-related disorders), and treatment for more than 111 days (median for all episodes) predicted dropout. Episodes that involved psychotropic drugs were more likely to end in dropout than those that did not. Psychiatrists were divided into two groups by the size (large or small) of the population per psychiatrist in private practice in their catchment area, and a large population predicted dropout.
Our study found that the diagnostic profile of patients treated by psychiatrists in private practice in Denmark is similar to profiles in other countries (4,5,6,7,8): the percentage of patients with affective disorders varies between 25% (7) and 46% (4), and the percentage with anxiety disorders varies between 11% (4) and 36% (8). In our study and those from other countries, female patients outnumbered male patients and most were between the ages of 30 and 50.
In a 1988 study of 166 patients treated by psychiatrists in private practice in Copenhagen, 63% had a diagnosis of neurosis, 8% had affective psychosis, and 8% had a personality disorder (2). The differences with our study may result from different diagnostic systems—ICD-8 in 1988 and ICD-10 in our study—although the patient population in Copenhagen may be different from that in the rest of the country.
Our study found that the three largest diagnostic groups (none of which includes psychosis) accounted for 79.1% of the episodes involving men and 86.1% of those involving women. In the 1988 study 71% of patients had a nonpsychotic diagnosis (63% neurosis and 8% personality disorder) (8). Thus our findings do not support the hypothesis that the number of patients with psychotic disorders treated by psychiatrists in private practice increased after decentralization and with the increasing numbers of psychiatrists in private practice. This is further underlined by the negligibly low proportion in our study of episodes with a diagnosis of a schizophrenia spectrum disorder—4.1% of episodes involving men and 2.5% of those involving women.
Having a personality disorder predicted dropout, which is similar to findings in the United Kingdom (13). Because younger age was also a strong predictor of dropout, special attention should be given to younger male patients in treatment for personality disorders because they are at higher risk of dropout. In areas a large number of patients per practicing psychiatrist, patients often must travel a long distance to see a psychiatrist, which may explain the higher likelihood of dropout among these patients—a finding also reported in Canada (14). Telephone reminders might be used to lower the dropout rate (15).
Our study has limitations. Although about 15% of all psychiatrists in private practice reported data to the database each year during the study period, they do not represent all psychiatrists in private practice in Denmark. For example, none of the psychiatrists reporting data were from the capital city of Copenhagen. However, all Danish psychiatrists have the same basic training and must complete the same specialization courses, a process overseen by the Danish National Board of Health. In addition the average age of patients in our study is similar to the average age of patients treated in private psychiatric practice in Denmark between 2003 and 2006 (47.5±15.1 and 45.0±15.4 years, respectively). Also, the percentage of male patients in our study was similar to the percentage in Danish private psychiatric practice (35.8% and 37.6%, respectively) (personal communication, Danish Regions 27, March 2009).
The database has received unrestricted research and educational grants from Novo Nordisk, SmithKline Beecham, GlaxoSmithKline, and Wyeth.
Dr. Munk-Jorgensen has received unrestricted educational grants or honoraria from H. Lundbeck A/S, AstraZeneca, Eli Lilly and Company, Janssen-Cilag A/S, and Bristol-Myers Squibb. Dr. Andersen has received support for the database, salary for the Web master, and payments for two meetings of the database group from Wyeth.