Abstract
Deliberate foreign body ingestion (DFBI) is defined as non-accidental ingestion of a true foreign body (non-nutritive items) for parasuicidal reasons. 1 It is an under-recognised, self-injurious behaviour that can be seen in patients with underlying mental disorders such as psychosis in response to hallucinations and delusions, in intellectual disabilities as pica and as malingering in institutionalised patients from prison or long stay mental health facilities. 2 It is frequently associated with severe personality disorders as an impulsive or premeditated repetitive behaviour that is resistant to intervention. 2
DFBI often occurs in clusters in a small population of patients, providing a challenge for healthcare providers and has a significant financial impact on the healthcare system. One observational study 3 has suggested an average cost of $11,000 US dollars per admission.
ESGE guidelines for foreign body removal
Emergent: 2 to 6 h, Urgent: within 24 h, non-urgent: within 72 h.
Identified risk factors for recurrent DFBI presentations include lower socioeconomic status, being a prisoner and male gender. 6 The available literature is relatively sparse with many case reports, case series 2 and consensus guidelines addressing the acute management and the endoscopic or surgical removal of the foreign bodies (FBs) but limited research on appropriate preventative measures, associated psychopathology, risk factors or potential psychological interventions.
Thus, we aimed to audit and analyse existing endoscopy databases at three large Australian teaching hospitals: To assess the characteristics of patients admitted with DFBI; report the types of foreign bodies ingested and evaluate the medical care and the outcomes of recurrent presentations.
Methods
Patient cohort characteristics
Adult patients with who presented with DFBI between January 2013 and September 2020 were identified from prospectively maintained endoscopic databases at three major teaching Australian hospitals (an urban and two outer Metropolitan hospitals) in two different states. Overlapping admissions to a 4th hospital were captured and included. Data were extracted by authors PK, WS and ZT from discharge summaries and progress notes via electronic medical records (eMRs) to include patients with clinically diagnosed mental health illness. Data collected included demographics; comorbidities including underlying mental health diagnosis, history of substance use disorder, history of sexual abuse; foreign body related data such as type and number of foreign bodies ingested, method and outcome of endoscopic retrieval; time to endoscopy; active mental health plan and hospital admissions–related data: number and time between presentations and length of stay (LOS). Where LOS was more than 48 h, we conducted a review of the progress notes to delineate the reason for extended admission. Patients with accidental foreign body ingestion or with no underlying mental illness were excluded from this study.
Statistical analysis
We calculated descriptive statistics for the cohort. Categorical data were presented as number and percentage. Continuous data were presented as mean and standard deviation (SD) and analysed using the independent samples
Ethics approval
The study protocol was reviewed and approved by the Central Adelaide Local Health human research ethics committee (2020/AU/1/5B3B312). Subsequent site-specific approvals were obtained for all four hospitals from respective Local Health Network research governance teams.
Results
Patient characteristics
Sample characteristics
Summary of comorbid mental health diagnoses
Hospital admissions
Most of DFBI admissions were under a gastroenterology unit (65%), followed by an upper gastrointestinal surgery unit (10.8%). DFBI-related admission remained variable across all hospital sites. Readmissions from patients with previous DFBI-related presentations contributed significantly to the overall increased trend of such admissions throughout the 8-year study period.
In this cohort, a small proportion of patients had significantly increased frequency of repeat presentations which contributed to the bulk of overall admissions (Figure 1). Of all the admissions, 51% were accounted for by 18/35 patients, 17 of which had underlying borderline personality disorder (BPD). Among the 18 patients who had recurrent admission, the mean number of days lapsed between consecutive DFBI-related admissions per patient was 59 days. The average length of stay (LOS) was 4.1 (SD 5.9) days with 44% of all admissions having a prolonged LOS of more than 48 h. Based conservatively on the National Hospital Cost Data Collection Report,
7
average cost per hospital day (AUD$2606), we estimate a total cost of 1,773,643.,60 AUD for our cohort. Histogram (frequency) of presentations per patient.
Based on discharge summaries and inpatient documentation, where reviewed, around half of the encounters (87/166) had a mental health review during the admission and half (89/166) had a documented community mental health action plan. The vast majority (151/166) of these patients were treated with an antidepressant and/or an antipsychotic at the time of the admission.
Reason for prolonged LOS (>=48 h)
Foreign bodies
A total of 109/166 (65%) admissions involved ingestion of ‘high risk objects’ such as button batteries, sharp objects and magnets with yet an increased trend of admissions over the years.
Type of FB ingested—summary table
Others: Ceramic, broken spectacles, leather bracelet, scalpel blade, lighter, marbles, watch clip.
Around 86% of patients had a plain X-ray on presentation, 4% had a CT scan and 10% had no imaging (mainly in repeated presenters).
In 76.5% of cases, an endoscopic intervention was performed to retrieve the FB, while no endoscopy was performed in the remaining patients as the FB was beyond the Pylorus on X-ray. All the endoscopies but one was done urgently (within 24 h of presentation) with a mean time to endoscopy of 9.32 (SD 6.19) hours.
There were three reported serious complications periendoscopic retrieval (2 duodenal perforations and 1 entero-enteric fistula), all due to delayed emergency presentations and absconding in patients with recurrent presentations. We note that the endoscopy was performed within 2 h of second presentation to remove razor blades, magnets and utensils.
Discussion
Our study shows that DFBI is an increasingly common problem in Australian hospitals. This appears to be similar to recently published data from the UK(8). A small proportion of patients with underlying mental disorders were responsible for 166 admissions over the period of the study. There was a slight female preponderance. Around 2/3 of included patients had a documented diagnosis of BPD, 95% of which were repeated presenters. Substance use disorder was present in 40%.
Hospital resource utilisation was significant due to a relatively prolonged LOS with 680 total inpatient days among this group during the study period; involving extensive nursing, medical, procedural and mental health support. This, along with an increase trend of DFBI over the years is concerning and warrants increased attention on preventative strategies and management and is likely to benefit from a multidisciplinary approach across medical and psychiatric specialties.8,9 Particularly, for repeated presenters, a clear shared management plan can help to provide consistency in responses and considering medico-legal risk. 10 The role of psychodynamic mechanisms in DFBI should be appreciated. 11 Developmental trauma is common and some authors have suggested that oral dependency and masochistic needs are fulfilled in the act of DFBI, which can also be conceptualised as a form of focal suicide. DFBI may also serve to regulate affect. 12 The management of counter-transference and provision of empathic care is critical to minimise challenging inpatient behaviour. 13 While repeated presenters with BPD are complex and may not engage well in treatment, there is level 1 evidence that psychotherapy such as dialectical behaviour therapy (DBT), cognitive behavioural therapy (CBT) or mentalisation based treatment (MBT) may be beneficial. 14 Pharmacotherapy is not specifically useful in BPD to reduce self-harm but is essential to treat comorbid psychopathology such as depression or psychosis. 15 It is important to note the substantial rates of substance use, psychosis, depression and anxiety disorders within this sample. These conditions should be optimally managed to reduce the risk of subsequent repeated self-harm. Thus, appropriate mental health assessment is critical during presentation to assess the need for further intervention and follow up planning. 10
Dangerous objects were the most frequently used: Sharp objects, Batteries and razor blades. When endoscopic retrieval of the FB was indicated, all procedures but one were performed within 24 h of presentation which aligns with the ESGE clinical guidelines for foreign body removal. Those guidelines are difficult to adhere to sometimes and perhaps not the right metric in the setting delayed presentation; hence, treatment should be individualised on a case-by-case basis.
The main reason for increased length of stay was psychiatric.
No deaths were noted and serious complications occurred only in 3 cases, all associated with delayed ED presentations for patients with high-risk behaviours such as absconding. Target time to endoscopy was achieved after second presentation; however, it was factually prolonged due to absconding and returning to the hospital.
In terms of limitations, data were extracted from discharge summaries and confirmed in medical records where possible. The accuracy of these diagnoses and potential missing data such as treatment with approved psychotherapy for patients with BPD needs to be considered. Similarly, community based mental healthcare plans may not have been well documented in hospital records. Future studies should seek to link hospital and community records.
In conclusion, DFBI in mental health patients is a common, recurring and challenging problem with safe endoscopic outcomes but likely significant financial impact on the healthcare system. It requires collaborative research to further guide a holistic patients management.
Footnotes
Disclosure
Funding
ORCID iDs
