Since the turn of the new century, many western countries have progressively implemented intensive home treatment (IHT) as an alternative to inpatient management for people experiencing a psychiatric acute episode. However, the profiles of patients more suitable for IHT have not been clearly defined yet,1 and research in this regard is scarce,1–4 mainly limited to the United Kingdom,2–4 and inconclusive. Therefore, there is an evident need for more international studies addressing this topic. In this vein, the aim of the present study was to evaluate, in a wide Spanish sample, which sociodemographic and clinical characteristics determine that a patient with an acute psychiatric illness is admitted to hospital or to IHT.
Material and methodsStudent's t-tests, chi-square tests, and association measures were used to compare the sociodemographic and clinical characteristics of the patients voluntarily admitted in-ward (N=216) or to IHT (N=90) during the period comprised between 1st February 2017 and 31st July 2018. A logistic regression model was built to assess the contribution of each factor, using a backward stepwise selection strategy (enter p value<0.05, remove p value>0.10). The discrimination power of the model was evaluated by means of the area under the receiver operating characteristic curve (AUC); sensitivity, specificity, and positive and negative predictive values were also calculated. Finally, a 5-fold cross-validation technique was used to test the performance of this model. All hypotheses were tested with a two-sided significance level of 0.05. Data were retrospectively obtained from the medical records of the emergency ward of Hospital de Sant Pau in Barcelona (the main admission gateway), except for severity at admission, which was obtained from the notes of the admission in hospital or IHT, and was assessed by means of the 17-item Hamilton Rating Scale for Depression, the Young Mania Rating Scale, and the Positive and Negative Symptom Scale. As IHT always requires patient's will, compulsory inward admission was considered an exclusion criterion.
ResultsPatients admitted inward were more likely to have previous suicide attempts, severe behavior disorders, active suicidal thoughts, or a non-psychotic/non-affective disorder. By contrast, patients admitted to IHT were more likely to have passive death wishes or a psychotic disorder (Table 1). Affective disorders were similarly distributed between both groups. There were no differences in duration of illness, presence of previous admissions, active substance use, level of insight, and clinical severity. Most patients had a partial insight and a moderate-to-high clinical severity. The regression model included age, having a caregiver, previous suicide attempts, non-psychotic/non-affective disorder, suicidal thoughts, and severe behavior disorder. The AUC of this model was 0.76 (95% CI=0.71–0.82). Sensitivity was 84.51%, specificity was 42.22%, positive predictive value was 77.59%, and negative predictive value was 53.52%. Results of the cross-validation process suggested that the real predictive power of the model was very low (pseudo-R2=0.14).
Sociodemographic and clinical characteristics of patients admitted to hospital (N=216) vs patients admitted to IHT (N=90).
Sociodemographic characteristics | Clinical characteristics | ||||||||
---|---|---|---|---|---|---|---|---|---|
Hospital | IHT | Hospital | IHT | ||||||
Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||
Age (years)* | 49.71 | 18.84 | 43.50 | 15.66 | Duration of illness (years) | 18.00 | 18.74 | 13.84 | 17.12 |
Hospital | IHT | Hospital | IHT | ||||||
---|---|---|---|---|---|---|---|---|---|
N | % | N | % | N | % | N | % | ||
Sex (female) | 119 | 55.09 | 52 | 57.78 | Previous admissions | 150 | 69.44 | 56 | 62.22 |
Educational level | Previous suicide attempt(s)* | 53 | 24.53 | 6 | 6.67 | ||||
No studies | 8 | 3.70 | 0 | 0.00 | Diagnosis* | ||||
Primary studies | 37 | 17.13 | 17 | 18.89 | Affective disorder | 89 | 41.20 | 36 | 40.00 |
Secondary studies | 73 | 33.80 | 31 | 34.44 | Psychotic disorder | 91 | 42.13 | 50 | 55.56 |
High education | 49 | 22.69 | 26 | 28.89 | Other disorders | 36 | 16.67 | 4 | 4.44 |
Unknown | 49 | 22.69 | 16 | 17.78 | Substance use | 36 | 16.67 | 13 | 14.44 |
Employment situation | Severe behavior disorder* | 51 | 23.61 | 9 | 10.00 | ||||
Active | 36 | 16.67 | 9 | 10.00 | Level of insight | ||||
Off sick/unemployed | 64 | 29.63 | 37 | 41.11 | Low | 49 | 22.79 | 15 | 16.67 |
Permanent disability | 62 | 28.70 | 30 | 33.33 | Moderate | 110 | 50.93 | 43 | 47.78 |
Pensioner | 41 | 18.98 | 8 | 8.89 | High | 57 | 26.51 | 32 | 35.56 |
Never worked | 12 | 5.56 | 6 | 6.67 | Clinical severity | ||||
Unknown | 1 | 0.46 | 0 | 0.00 | Minimal | 38 | 17.59 | 16 | 17.77 |
Caregiver* | 137 | 63.43 | 74 | 82.22 | Mild | 56 | 25.93 | 28 | 31.11 |
Moderate | 62 | 28.70 | 24 | 26.67 | |||||
Severe | 60 | 27.78 | 22 | 24.44 | |||||
Suicidal behavior | |||||||||
Absence | 106 | 49.07 | 59 | 65.56 | |||||
Passive death wishes | 29 | 13.43 | 19 | 21.11 | |||||
Suicidal thoughts | 56 | 25.93 | 6 | 6.67 | |||||
Suicide attempt | 25 | 11.57 | 6 | 6.67 |
Both severe behavior disorders and active suicide thoughts seem logical contraindications for IHT, and have been also associated to hospital admission in previous studies.2,4 Similarly, history of previous suicide attempts may be related to inpatient management because of its relationship with a higher suicidal risk.5 However, which disorders are more frequently managed at home remains unclear. Our results show that having a psychotic disorder increases the probability to be admitted to IHT, while having a non-psychotic/non-affective mental disorder increases the probability of in-ward treatment, which is consistent only with some previous research.1,2,6 These discrepancies between studies may lie in the different distribution of mental health services among areas.
Despite these results, the low predictive power of the regression model suggests that the decision of offering an in-ward or a home admission may also depend on other variables than those directly associated with the patient, like the availability of beds in each facility or the psychiatrist's personal preferences, as suggested elsewhere.1
Last but not least, it is interesting to emphasize that both groups had a similar distribution of the rest of sociodemographic and clinical variables, including clinical severity and level of insight. In our opinion, this supports that IHT can be a real alternative to classical hospitalization.
In conclusion, IHT may be able to take on the management of most of the patients suffering a psychiatric acute process who accept treatment, keeping the hospital only for those situations in which there is a risk for the physical integrity of the patient or others. So, IHT may be a cost-effective7 way to reinforce care outside hospitals, as proposed by some authors in the context of the COVID-19 pandemic.8,9 Further studies with experimental designs may help to draw more robust conclusions on which factors are associated with a successful home-treatment.
Conflict of interestThe authors declare no conflicts of interest.