The Importance Of A Consultation Liaison Psychiatry Service In General Hospitals: A Brazilian 6-Years Experience

Background: The importance of consultation-liaison psychiatry (CLP) services have been progressively growing in all general hospitals. Methods: Data were collected from medical records of patients treated by the CLP service, from 2014 to 2020, including clinical-demographic data, characteristics of hospitalization, reason for request, treatment plan, and clinical outcome. For these analyses, the chi-square test, z-test, and Kruskal-Wallis test were used. The level of signicance was set at 5%. Results: A total of 2,742 consultation requests were evaluated by the psychiatry team. The number of requests made grew by 109.64% over the years. In 58.21% of cases, the patient was in the ward, and 33% were requested by Internal Medicine. There was a statistically signicant difference (p <0.001) in the distribution of reasons for request according to patients' characterization variables. Conclusions: This study sought to analyze the functioning and growth of CLP in a tertiary Brazilian general hospital. The present ndings reinforce the importance of this service and the need to implement and develop CLP in general hospitals.


Introduction
The Consultation-Liaison Psychiatry (CLP) service was developed based on the need for interaction and integration of the mental health professionals with other medical specialists in the general hospital; in order to reduce the social stigma caused by the psychiatric disorders and improve the care for the psychiatric patients based on the biopsychosocial model [1][2][3][4]. Past research has already observed that CLP services are cost-effective as they decrease the length of hospitalization stay [5][6][7]. In addition, CLP services expand the detection and interventions on psychiatric comorbidities, with a good acceptability by patients [6]. In addition, considering the CLP service as an area of increasing importance in the rearrangement of mental health services, it is essential that their demand, service pro le, and patients' characteristics can be taken in account, in order to detect obstacles and limitations of the service, as well as improve the cost-bene t ratio of clinical hospitalizations. Based on these premises, the aim of this study was to outline the pro le of a CLP service of a large tertiary Brazilian general hospital.

Materials And Methods
The present investigation is an observational, cross-sectional, retrospective and descriptive study of the CLP service at a tertiary care university-a liated teaching hospital, Hospital de Base, located in the city of Sao Jose do Rio Preto, Brazil. The Hospital de Base, located in the city of Sao Jose do Rio Preto and connected to the FAMERP, is one of the largest teaching hospitals in Brazil, with 146,467 visits/year, 89,754 attendances in emergency, and 41,226 hospitalizations. The hospital has a CLP service with expertise in psychiatric consulting and production of scienti c articles [8].
All data were collected retrospectively from 2,742 inpatient consultation requests from medical records using a structured form (Google Forms). The collection was made independently through the data obtained from the consultants' requests and responses.
This study was approved by the Research Ethics Committee of the Faculdade de Medicina de Sao Jose do Rio Preto (FAMERPnumber 36147620.0.0000.5415).

Statistical analysis
The variables analyzed were number of requests for the CLP service; hospital process variables (lag time, length and number of hospitalizations), sociodemographic pro le, clinical and psychiatric diagnosis (reason for requesting, previous psychiatric history, previous/diagnosis/hospitalization, and psychiatric family history), interventions (pharmacological and non-pharmacological treatment plan, prescribed drugs, reassessment and non-evaluation).
Statistical analyses were performed using SPSS 20 software. For descriptive analyses, the chi-square test was used; and z-test was used to compare the proportions of the columns. The Kruskal-Wallis test was used to compare the response time after a request for consultation between the main requesting areas (emergency, in rmary and ICU). The level of signi cance was set at 5%.

Results
A total of 2,742 consultation requests for the CLP service were evaluated, which correspond to 0.73 to 1.50% of the total hospital admissions in the period. The number of CLP requests grew by 109.64% in the period of the study (Figure 1). Regarding patients' sociodemographic characteristics, 1,516 (55%) were female, the mean age was 43.22±18.00 years and the median age was 43.00 years; 908 (33.2%) had not completed elementary school, 1,015 (37.5%) were married or in a stable relationship; 762 (27.8%) were Catholic, 1,850 (67.47%) were living with a family member; and 1,725 (62.6%) did not work due to retirement (n=479, 26.9%), unemployment (n=410, 23%), due to health conditions (n=244, 12.9%), or being a "housewife" (n=173, 9.7%). The majority (73.60%) had a psychiatric history, 55.43% had a previous psychiatric diagnosis, 29.54% had a psychiatric follow-up, 22.25% had a previous psychiatric hospitalization, and 52% had a family psychiatric history.
Requests for CLP service were made 8.20 ± 20.60 days after the patient's admission, and the team's responses were provided after 9.65 ± 21.85 days. The response time between the request and response was of 1.43 ± 4.36 days. Regarding the evaluation site, 1,596 (58.21%) patients were in the ward, 949 (34.61%) in the emergency room, 182 (6.64%) in the ICU, 3 (0.11%) in the operating room, 2 (0.07%) in the ambulatory room, 1 (0.04%) unavailable, 7 (0.26%) not informed, and 2 (0.07%) in other situations. Most consultations (n=901, 32.5%) were requested by Internal Medicine. The main reasons for requesting CLP evaluation were: a) presence of a previous psychiatric disorder; b) use of psychiatric medication; c) presence of ill-de ned symptoms; and d) history of suicide attempts. Reassessments were necessary for 415 patients (16%), and the CLP service did not evaluate 196 patients (7.2%) because of reasons such as discharge, inability to answer correctly, coma/intubation, evasion, and death.
Among the pharmacological treatment plans, the introduction of psychotropic drugs was the most commonly performed procedure, occurring in 894 patients (30.86%). The most frequently prescribed classes of drugs were, in order, antidepressants, antipsychotics, and benzodiazepines. In addition, outpatient follow-up was the CLP service's most frequent recommendation to continue the treatment, followed by being admitted to a general hospital, referral to a psychiatric emergency, and referral to one of the services of the publicly psychiatric care network.
There was a statistically signi cant difference in the distribution of the reasons for CLP evaluation according to the patients' characterization variables: gender (p≤0.001), city (p=0.018), presence of a companion during hospitalization (p≤0.001), age (p≤0.001), and marital status (p≤0.001), as shown in Tables 1 and 2. Considering the three main hospital scenarios (emergency, in rmary, and ICU) that requested consultation, the response time was shorter for the emergency (0.92±0.83), and comparable for the in rmary department (1.46 ± 1.17) and ICU (1.53 ± 1.24) (p <0.001).
There was a statistically signi cant difference when the consultations were not performed, and the ICU presented a higher number of non-assessments (p≤0.001). Regarding the reasons for non-evaluation, it is noteworthy that "Hospital discharge" was less common in the ICU, "Absence due to laboratory test/procedures" was less common in the Emergency department and more common in the In rmary department, "Evasion" was more common in the Emergency department, and "Impossibility to respond (coma, intubation)" was more common in the ICU (p≤0.001). The reason for not evaluating due to "Inability to respond correctly" and the other reasons were not associated with the requesting department (Table 3).

Discussion
Since its introduction in Brazil in 1977, the CLP service has been the link between patients, family, medical staff, and external health services in the eld of mental health. In many cases, it is also the rst opportunity for psychiatric care, further emphasizing the importance of this service [3,9,10,11]. CLP services are present in 86% of general hospitals, with the vast majority in university hospitals [9,[12][13][14]. To the best of our knowledge, this is one of the largest studies on Brazilian CLP services, which involved an elevated number of requests (2,742), and with a large time interval (six years).
The number of CLP requests corresponded to 0.73 to 1.50% of the total hospital admissions in the period. Studies with different follow-up periods have disclosed that CLP requests correspond to 0.9-2% of all hospital admissions [13][14][15][16][17][18][19][20][21][22]. Despite the similarity to literature, this frequency remains below what is expected from the prevalence of mental disorders in Brazil [21,23]. One reason for this discrepancy is probably the passive screening by the CLP teams [24].
The sociodemographic characteristics observed in this study agreed with other published reports in the eld [15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31]. Similarly to other long-term studies, the geriatric population corresponded to almost 10% of the sample [15,16]. Another study conducted at the Hospital de Base of Sao Jose do Rio Preto from 2010 to 2014 showed that 14.33% of the consultations were made for elderly patients, which is the same proportion as our data [32]. In contrast, Bambarén [25] found 41.6% total psychiatric consultations for the elderly population in Rio Grande do Sul, Brazil and Wancata [33], with 45.3% in a hospital in Vienna, Austria. The di culty in managing these patients, since they may have greater morbidity, comorbidities, and drug interactions, may be obstacles to psychiatric consultation in the coming years.
The CLP response time was in agreement with other studies, with a large portion being answered before 48h [16]. In this study, a faster response was observed for the Emergency sector, due to the serious nature of the cases (p<0.001). In addition, the time between admission and request for consultation (lag time) was also consistent with other studies, varying between 3 and 15 days [13,15,19]. In a study by Nakabayashi [19], endocrine, metabolic, and nutritional conditions were observed and early consultations were requested, due to dysfunctional conditions that required rapid intervention. In cases where there was no possibility of evaluation, it was observed in this study that such cases occurred in the ICU (p<0.001). As the main change was "intubation", a better communication between the teams seems to be essential. Such data have not been frequently related in other studies in this eld; this sheds light on this very common occurrence in general hospitals.
Regarding the frequency of the teams that requested CLP evaluation, Internal Medicine has been the most frequent [13,15,16,19,21,[28][29][30]. This nding would be associated with the possible greater sensitivity and accuracy of clinicians in detecting behavioral alterations. Furthermore, according to Kishi et al. [29], clinical physicians have higher measures of empathy than those in specialties focused on technology, such as anesthesiology, surgery, and other surgical subspecialties, which may explain the reason for delays or lower demands on surgical wards.In addition, the in rmary was the location of more than half of the cases of CLP requests. This nding was discordant to what has been found in other studies; Huyse [34] observed that the emergency corresponded to 33% of the service, mainly due to suicide attempts, self-mutilation, and intoxication by psychoactive substances [33,34].
As for the therapeutic plan, the majority have focused on drug treatment, as well as other research [16]. Antidepressants have been selected in most cases in several studies [21,25,32]. After discharge, the patient was referred to a service for follow-up if necessary. In a study by Rigatelli [10], psychiatric outpatient clinics were the main destination (29%), followed by primary care units (27%), private practices (8%), and maintenance of hospitalization (6%).
Most studies in the eld of CLP seek to show the cost savings of this type of service, but there are important methodological aws, the impossibility of building an adequate control group, short follow-up time, absence of prospective data, and standardized documentation [9,15,37]. This absence of quality indicators conveys the erroneous feeling that the CLP service does not produce anything or does not have concrete goals [38]. Therefore, as long as randomized clinical trials are not carried out, it is important to complement them with retrospective/prospective studies to record the evolution and development of the CLP service [16,39]. Moreover, due to the low percentage of requests compared with total hospital admissions, the CLP has been mostly seen as a reactive measure by general hospitals to obtain emergency psychiatric care, rather than as a mental health strategy. Therefore, CLP services often suffer from exclusive issues such as stigma, nancial disincentives, lack of an adequate multidisciplinary team, or low remuneration [38, [40][41][42]. This is re ected in the high turnover of the CLP team, often with a short experience and less production of research data [13,27,43,44].
The present study has several limitations. Despite being a relatively long-term study, as it is transversal and retrospective, it does not re ect the national reality. Different hospital con gurations, consultants, study periods, sample sizes, and diagnostic classi cation systems can cause heterogeneity with other services. Standardized psychiatric scales and structured clinical interviews were not used in this study. Due to the methodology, the patients' hospitalization time and other interesting data, such as adherence by the team, were not obtained.