Routine investigations for patients with mental and behavioural disturbances

Background The process of medical clearance aims to exclude a general medical condition as an underlying cause for mental and behavioural disorders and involves routine screening with special investigations. Current evidence, however, suggests that clinician gestalt should guide the need for special investigations and that there is no benefit to routine screening. Aim This study aimed to determine the effectiveness of and adherence to the Western Cape (WC) provincial guidelines for routine investigations of adult patients with behavioural disturbances. Setting This study was conducted at Mitchells Plain Hospital in Cape Town, South Africa. Methods This descriptive study was conducted at Mitchells Plain Hospital in Cape Town, South Africa. Data were collected from existing electronic registries over a 6-month period. Adult mental healthcare users were risk stratified into the probability of having a general medical condition and the results of their special investigations were described against their outcome. Results Of the 688 patients included in this study, 66% had abnormal vital signs and of the 312 patients who received special investigations, 56% were abnormal, including 18% who were clinically significantly abnormal. Abnormal special investigations changed the clinical outcome for 3 (<1%) patients. Adherence to the guidelines was reasonable (82%) but non-adherence resulted in unnecessary investigations. Conclusion The results of this study support the existing evidence that clinical assessment and clinician gestalt should guide the need for special investigations and that there is no benefit to routine screening in the emergency centre (EC). The results also demonstrate that non-adherence rarely changed patient outcomes. Contribution This study provides information on the value of routine screening investigations in ECs.


Read online:
Scan this QR code with your smart phone or mobile device to read online.

Introduction
Mental and behavioural disorders are one of the leading causes of years lived with disability (YLDs) worldwide. 1The vast majority of this burden (80%) occurs in low-and middle-income countries (LMICs) where nearly 20% of health-related disabilities are attributable to mood, psychotic and substance abuse disorders. 2In South Africa, the 12-month prevalence of mental health disorders is 16.5%, with an estimated lifetime prevalence of 30.3%. 3 The reported prevalence in the Western Cape province is the highest nationally with a 12-month and lifetime prevalence of 39.4%. 3,4Despite this significant burden, mental and behavioural disorders continue to be a low priority with regard to resource allocation. 5,6,7Previously (1980), this process was reported to add significant value with 34% to 46% of patients with mental and behavioural disorders found to have underlying medical conditions. 11However, more recent evidence suggests that appropriate special investigations should rather be guided by clinical history, vital signs and physical examination and not performed routinely. 12,13,14Patients with a primary mental and behavioural complaint with normal vital signs and a normal physical exam have less than 1% chance of having a clinically significant abnormal special investigation. 15,16A local study conducted at a district-level hospital by Credé et al. demonstrated that routine laboratory screening provides no additional information Background: The process of medical clearance aims to exclude a general medical condition as an underlying cause for mental and behavioural disorders and involves routine screening with special investigations.Current evidence, however, suggests that clinician gestalt should guide the need for special investigations and that there is no benefit to routine screening.

Routine investigations for patients with mental and behavioural disturbances
Read online: Scan this QR code with your smart phone or mobile device to read online.
in the assessment of patients with mental and behavioural disturbances on presentation at the emergency centre (EC). 8,9,17,18ank et al. performed a before-and-after study at a tertiary hospital where an institutional policy to eliminate routine special investigations for all patients with mental and behavioural disorders was implemented. 19Special investigations were subsequently performed at the discretion of the treating clinician, which drastically reduced EC length of stay by a mean of 5 h 30 min, reduced total hospital length of stay by 16 h and decreased the EC laboratory costs per patient by 78%.More importantly, there was no significant increase in adverse events or number of medical referrals and no deaths were reported. 19This approach is supported by local and international evidence, as well as professional bodies such as the American College of Emergency Physicians (ACEP) and American Psychiatric Association (APA). 13,17,20,21,22,23,24utine screening comes at a cost, adding to the financial burden, staffing requirements, risk of needle stick injuries and patient discomfort, as well as contributing significantly to EC crowding and staff and patient safety.'Crowding exists when there is no space left to meet the timely needs of the next patient requiring emergency care'. 25Crowding in the EC negatively affects quality of care, increases medical error and increases morbidity and mortality of all patients in the EC. 25,26ith prolonged EC length of stay, the risk to patient and staff safety also increases, as patients with mental and behavioural disturbances are often difficult to contain in a chaotic EC environment.The risk-benefit ratio between appropriate screening practices and the risks associated with EC crowding and staff safety should always be considered.
Mitchells Plain District Hospital's EC follows the Western Cape provincial 'guideline for routine investigations for exclusion of GMCs in patients presenting with mental and behavioural disturbances'. 27The guideline is aimed at reducing length of stay and categorises patients into three groups: (1) patients at high risk for a GMC; (2) known mental healthcare users (MHCU) who are not at high risk for a GMC; and (3) index MHCUs who are not at high risk for a GMC, based on their clinical features.According to the guideline, special investigations are then performed according to the patient's risk-category.Index MHCUs require a predefined set of special investigations, and MHCUs at high risk of a GMC require a predefined set of investigations as well as additional tests that are clinically indicated. 17,27Currently, there are no data that describe adherence to and effectiveness of the given guidelines.The data from Credé et al. was sourced more than 10 years ago and the mental health disease burden and community profiles have changed significantly since then. 4The aim of this study was therefore to assess the adherence to and effectiveness of the Western Cape provincial guideline for routine investigations of adult patients with mental and behavioural disturbances presenting to a districtlevel EC.

Study design
This study is a descriptive analysis and data were collected retrospectively from existing databases and registeries.

Study setting
This study was conducted at Mitchells Plain Hospital (MPH) in Cape Town, South Africa, which is in the Mitchells Plain Health District of the Western Cape province Metro Region.The hospital is 32 km from Cape Town's central business district and provides services to an estimated population of 600 000. 28It serves low-to middle-income communities of Mitchells Plain and mainly low-income communities of Philippi: A large nearby informal settlement.The MPH is a district-level hospital that has a psychiatric department that is headed by a single psychiatrist and has both a male and female ward on site.

Inclusion criteria
All adult patients (18 years and older) with mental and behavioural disturbances who presented to Mitchells Plain Hospital's EC between July 2019 and December 2019 were eligible for inclusion.For the purpose of this study, the definition of 'mental and behavioural disturbances' included any diagnosis from the World Health Organization's (WHO) international classification of diseases (ICD-10) Version: 2019 Chapter V: Mental and Behavioural Disorders (F00-F99).

Exclusion criteria
Patients who were admitted directly to the psychiatric department and bypassed the EC were excluded as well as those who were discharged directly from the EC, transferred to a different hospital, or directly admitted to another discipline.Patients who were clinically delirious with a clear cause (GMC) and referred to the appropriate discipline were not included in this study as the guideline does not apply to them.

Data collection and management
Data were collected in three phases according to the principles outlined by Gilbert et al. and Lowenstein et al. to minimise inconsistencies. 29,30In Phase I, a retrospective review of the data of all included patients was performed.Data were exported from the electronic registry, Hospital and Emergency Centre Tracking and Information System (HECTIS) along with patients' demographic details, their disposition category, folder numbers and initial vital signs on triage.An ICD-10 search from within the database was carried out by the database manager and de-identified data were exported to a spreadsheet.Folder numbers were used to track patients through all the phases of data collection.The notes were scrutinised, and patients were grouped into three categories according to their risk of GMC as a cause for their symptoms of mental and behavioural disturbance as per the Western Cape (WC) guideline.The South African Triage Scale (SATS) was used to group patients into triage categories (colour coded as green, yellow, orange and red) based on their presenting complaints, vital signs, clinical discriminators and severity of illness on presentation to the EC. 31 An adult MHCU who is aggressive on presentation is triaged orange according to SATS, indicating that such a patient needs very urgent care. 31In this study, the following thresholds were used for normal vital signs: systolic blood pressure: 90 mmHg -139 mmHg, diastolic blood pressure: 60 mmHg -89 mmHg, heart rate: 60 bpm to 100 bpm, respiratory rate: 12 bpm -20 bpm, oxygen saturation: 94% -100% and temperature: 35°C -37.5°C. 32The electronic records were primarily reviewed by two investigators, and a third investigator randomly cross checked ~5% of the case files to ensure that data collection was accurate.
In Phase III of the data collection process, the results of all special investigations performed on admission by the EC were obtained from the National Health Laboratory Services (NHLS) database and were classified into normal, clinically not significantly abnormal, or clinically significantly abnormal.This was also performed for results of special investigations carried out within 48 h of admission by the psychiatric department to ensure that patients with clinically significantly abnormal results, who were not otherwise investigated by EC, were not missed.Clinically significantly abnormal values were based on evidence and international consensus and defined thresholds for what is considered likely to contribute towards psychotic symptoms. 13,17,33,34,35able 1 summarises the reference values that were used.
Results of special investigations performed by both EC and the psychiatric department were assessed to determine whether they changed patients' outcomes.A change in outcome of patients was defined as the need for patients to be transferred to the medical or surgical department from the psychiatric ward.In addition, whether or not psychiatric inpatients received medical consultations (co-management) as a psychiatric inpatient, was also documented.
The effectiveness of the guideline was defined as how well it could distinguish between patients with mental and behavioural disorders as a result of a GMC and MHCU requiring psychiatric admission.To determine adherence, the proportion of patients who had special investigations according to the guidelines was calculated and expressed as a percentage per risk category.

Data analysis
All variables were described with summary statistics.Categorical variables were described as proportions or percentages and tabulated as necessary.The central tendency of continuous variables was described as medians, and quartiles were used to indicate spread.Tabulated variables were expressed as row percentages when outcome variables are grouped in columns and column percentages were used when descriptive variables are grouped in columns.Data were analysed using Statistical Package for Social Sciences (SPSS) Statistics Version 27.0 (IBM Corp. Released 2019; Armonk, NY: IBM Corp.).

Ethical considerations
Even though this study involved a vulnerable population, it posed a very low risk to patients as data were de-identified to protect patient identities.The data collection procedures only involved a folder and database review and no patient intervention or interaction occurred.Ethical approval was granted by the Human Research Ethics Committee of the University of Cape Town: HREC REF:678/2020 and institutional approval was obtained from MPH via the National Health Research Database: WC_202012_014.

Results
A total of 19,162 adult patients presented to MPH EC during the study period of which 960 (5%) were eligible for inclusion into the study.A sample of 688 (72%) were included, and Figure 1 provides a breakdown of the 272 (28%) exclusions.Among the excluded patients are 17 (2%) who had a clear diagnosis of a delirium for various reasons and were subsequently referred to relevant inpatient departments for admission.
Table 2 provides demographical and clinical details of the sample for each admission category.The sample had a strong male preponderance 455 (66%) and nearly 551 (80%) of all patients were younger than 45 years old.Based on their clinical details, 470 (68%) were known mental health users with no features of a GMC, 86 (13%) were index MHCU with no high-risk features of a GMC, and 132 (19%) had clinical features that increased their risk of having a GMC.The ratio Nearly all (99%) of the patients were triaged as orange, and the majority (68%) were known MHCU.Of those with highrisk features of a GMC, four (3%) were triaged red.MHCU with schizophrenia, schizotypal and delusional disorders contributed the most to the psychiatric burden with 491 (71%) of all admissions.Patients who required special investigations had a longer EC length of stay (7.5 h per patient), contributing to EC crowding -this includes all index patients and those at high risk for GMC (218 patients).The guideline correctly categorised the three patients who were transferred from the psychiatric ward to medical  http://www.sajpsychiatry.orgOpen Access department as high-risk patients.However, the use of the guideline could not change the patients' management plan in the EC.The three patients who were transferred to the medical department had the following reasons: Patient 1 was a 41-year-old female who had an abnormal TPHA and RPR -the institutional policy is that MHCUs are referred before these results are available as the turnaround time for TPHA results is long; Patient 2 was a 57-year-old female who had a clinically significant Vitamin B12 deficiency (not screened routinely in the EC on admission); and Patient 3 was a 76-year-old male patient whose EC laboratory results were normal; he developed seizures in the psychiatric ward and subsequently had an abnormal CT brain scan.

Discussion
This study demonstrates an 82% adherence to the provincial guidelines for routine investigations for exclusion of GMCs as a cause for mental and behavioural disturbances.The findings suggest that the application of the guideline did not significantly change patients' outcomes as decisions were based on clinician gestalt instead of results of special investigations, which is consistent with findings of other studies assessing the use of laboratory investigations for medical clearance in patients with mental and behavioural disturbances. 12,13,14,22Even though abnormal vital signs and abnormal special investigations were prevalent, they rarely resulted in a change in outcome as only 3 (<1%) patients, out of 312 patients who received special investigations, had a change in outcome.Although adherence was reasonable, non-adherence resulted in financial waste (>100 patients receiving unnecessary investigations) with no change in patient outcomes.
The high proportion of MHCUs with abnormal vital signs (66%) was surprising, considering the fact that abnormal vital signs could be considered a high-risk criterion to predict a GMC. 32Despite its prevalence, abnormal vital signs rarely changed the outcome as only 1 (0.2%) MHCU out of those who had abnormal vital signs, had a change in outcome (medical transfer from psychiatric ward).Variation in vital signs is common and evidence suggests that using groups or combinations of abnormal vital signs, together with clinician gestalt, is superior to one abnormal vital sign in isolation. 36he single patient who was transferred to the medical department had an abnormal heart rate and diastolic blood pressure.Further research should aim to assess which abnormal vital signs or combinations of abnormal vital signs have clinical value to predict undesirable outcomes in MHCUs.
Abnormal special investigations were prevalent (56% of patients who received special investigations) and included several clinically significantly abnormal results (18% of patients who received special investigations).Despite the high prevalence, it resulted in only three patients being transferred to the medical department from the psychiatric ward.Of the 56 patients (18%) who had clinically significant abnormal results, only 1 patient was referred to the medical department, despite there being 14 patients with clinically significant abnormal white cell counts and 14 with clinically significant abnormal thyroid stimulating hormone levels.Asymptomatic leucocytosis in patients seeking emergency care is fairly common and often transient; however, when it is coupled with clinical signs of a potential GMC, it has much more value to predict serious pathology. 37The fact that these MHCUs were declared medically fit and referred to the psychiatric department despite the abnormal special investigations questions the value of routine screening.The disposition of MHCUs was therefore based on clinician gestalt and not on the results of the investigations, which is a practice supported by an existing body of evidence that demonstrates that testing beyond what is clinically indicated for medical clearance in MHCUs rarely changes clinical care. 12,13,20,38,39The transfer of three patients from the psychiatric department to the medical ward did not occur as a result of a failure of clinical assessment or a lack of screening investigations.
Abnormal special investigations often resulted in MHCUs receiving medical consults as a psychiatric inpatient (comanagement), which is not surprising, as comorbid conditions are prevalent in MHCUs and many medical conditions can be safely managed while admitted in the psychiatric ward. 40This questions the utility and costeffectiveness of performing screening investigations in the EC as opposed to the psychiatric ward, considering that GMC as a cause for mental and behavioural conditions can be screened for reliably and safely with a thorough clinical assessment and clinician gestalt. 12,22From a risk-benefit ratio perspective, patients who wait for special investigations in the EC have a much longer EC length of stay, which increases crowding and affects staff and patient safety.In our sample, none of the index presenters' special investigations affected their outcome and none required medical transfer, despite the presence of abnormal results; however, each MHCU stayed for nearly 8 h longer in the EC.This makes a strong argument that medical workup that does not affect patient outcomes should ideally be performed in the psychiatric ward, especially if a medical department is available for consultations, to decrease EC crowding and allow patients to get to definitive care as soon as possible.
No previous studies have assessed effectiveness and adherence to the Western Cape guideline for exclusion of GMCs in adult patients presenting with mental and behavioural disturbances.Adherence to the provincial guidelines was reasonable (82%) but non-adherence resulted in financial waste (>100 patients receiving unnecessary investigations) with no change in patient outcomes.Reasons for non-adherence were not explored in this study and the financial impact was not quantified.Additional special investigations performed within the first 48 h as psychiatric inpatients also did not result in change in clinical outcomes, further strengthening the argument against routine screening.

Limitations
This study sample involved a single facility and only included data over 6 months.With regard to external validity, the authors are of the opinion that results are generalisable, because medical clearance in the public sector is standardised and follows the same provincial guideline.The findings of this study depend on clinical gestalt and therefore need to be contextualised: considering the single centre setting, it is reasonable to expect similar results in other emergency physician run ECs.The study results are particularly valid for facilities with both medical and psychiatric departments on site (in the same facility) but may not be applicable to psychiatric hospitals with no medical department (e.g.specialist psychiatric hospitals).Even though data on other investigations were collected (including drug levels, Vitamin B12, urine analysis, pregnancy tests, HIV viral loads and CD4 counts, creatine kinase, drug tests, etc.), it was not reported on as this project focused on routine investigations only, according to the provincial guideline.
Future studies should prospectively investigate the impact of no routine testing on patient outcomes, EC crowding, hospital expenditure and length of stay.Future studies should also involve multiple facilities and include speciality hospitals.

Conclusion
The results of this study support the existing evidence that clinician gestalt should guide the need for special investigations and that there is no benefit to routine screening in the EC.The results also demonstrate reasonable adherence to the current guidelines even though this rarely affected patient outcomes.Decisions that did change patient outcomes were based on clinical findings and clinician gestalt, not abnormal special investigations, or vital signs, despite these being prevalent.Guidelines, like these should ideally undergo an impact assessment before they are adopted.A potential benefit for one patient may impact the outcomes of other patients indirectly, especially in ECs where crowding and increased length of stay are linked with poor outcomes.The level of emergency care provision should also be considered as guidelines or clinical decision rules rarely trump clinical gestalt in (academic) settings where emergency medicine is practised.In this setting, clinicians chose clinical gestalt over the guideline on numerous occasions, which probably reflects the effectiveness of the guideline.The argument for routine screening with special investigations to occur in ECs is strongly challenged as the risk of the effects on crowding, as well as staff and patient safety, outweighs the lack of benefit from routine special investigations in addition to a thorough clinical examination and clinician gestalt.

TABLE 1 :
Reference values for laboratory investigations.
17urce: Adopted from Credé A, Geduld H, Wallis L. Assessment of routine laboratory screening of adult psychiatric patients presenting to an emergency centre in Cape Town.S Afr Med J. 2011;101(12):891-89417TPHA, rapid plasma reagin http://www.sajpsychiatry.orgOpen Access of 2:1 male predominance extended to the categories where no high-risk features for GMC were present, while nearly 50% (60 out of 132) of all MHCUs with high-risk features of a GMC were female.The highest proportion of MHCUs at high risk of a GMC and index presenters were in the 18-25-yearold category while the highest proportion of known MHCUs were in the 26-35-year-old category.

Table 3
depicts all abnormal vital signs for each admission category with subsequent psychiatric department outcomes.A total of 455 (66%) of all the patients included in the final sample had one or more abnormal vital sign.The most prevalent abnormal vital signs were diastolic blood pressure (32%), heart rate (29%) and systolic blood pressure (28%).

TABLE 2 :
31mographical and clinical details for each admission category: n (column%).Triage categories were based on the SATS.31 Note: Percentages may not add to 100% because of rounding.GMC, general medical condition; MHCU, mental healthcare user; IQR, inter quartile range; EC, emergency centre.*, none required transfer to the medical department and was co-managed (by internal medicine and psychiatry) as psychiatric inpatients.The only patient with clinically significantly abnormal results who had a change in outcome was a MHCU with neurosyphilis.

Table 5
consultations as a psychiatric inpatient.No clinically significantly abnormal sodium results were present but of the clinically significantly abnormal creatinine, white cell count and thyroid stimulating hormone results, none required transfer to the medical department and was co-managed (by internal medicine and psychiatry) as psychiatric inpatients.EC, emergency centre; GMC, general medical condition; ICD-10, international classification of diseases.

TABLE 3 :
Vital signs on presentation for each emergency centre admission category with subsequent psychiatric department outcomes.
Percentages may not add to 100% because of rounding; GMC, general medical condition; POC point of care; MHCU, mental healthcare users.*, Haemoglobin (g/dL) performed in only 583 (85%) patients.http://www.sajpsychiatry.orgOpen Access Table 6 depicts a summary of special investigations and adherence to guidelines for each EC admission category and the subsequent psychiatric department outcomes.The EC adhered in 91% (78/86) of the index presenter category, 77% (361/470) of the known mental health user category and 95% (125/132) of the high-risk of GMC category.The overall adherence was 82%.Non-adherence did not affect the outcomes for patients in the index presenter and high risk of GMC categories but resulted in 109 patients receiving unnecessary special investigations (not indicated) in the known mental healthcare user category.

TABLE 5 :
Abnormal special investigations performed by the psychiatric department within 48 h of admission for each emergency centre admission category with subsequent psychiatric department outcomes.

TABLE 4 :
Abnormal special investigations on presentation for each emergency centre admission category with subsequent psychiatric department outcomes.

TABLE 6 :
Summary of special investigations and adherence to guidelines for each emergency centre admission category and subsequent psychiatric department outcomes. http://www.sajpsychiatry.orgOpenAccess