Factors affecting discharge against medical advice (DAMA) at a private level 1 trauma centre: A retrospective cohort study

Materials & Methods

This retrospective cohort study was undertaken at a private trauma centre with level 1 capabilities, catering to the population of 3–4 districts near the border of the southern States of Tamil Nadu and Andhra Pradesh in India. Data was collected from a prospectively maintained trauma registry over 18 months, from March 2022 to August 2023. Data for all patients seen during the period in the Emergency department of Christian Medical College, Vellore, Tamil Nadu were retrieved. The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki (2013).

DAMA patients were compared with those who received definitive care for basic demographic details, mechanism of injury, injury severity scores (ISS), transfusion requirements, and vital signs. In participants with multiple injuries, a detailed chart review was done. All study participants were designated a predominant injury based on the region with the highest Abbreviated Injury Score (AIS). The patients were thus designated as having a predominant head injury, extremity injury, torso injury (neck to pelvis), isolated soft tissue injury, maxillofacial injury, vertebral and spinal cord injury (SCI), and ophthalmic injuries. Participants with an AIS score of >2 in two or more regions were classified as polytrauma. For DAMA patients, the reasons for the same were captured in the trauma registry and divided into four main categories: (i) financial constraints (inability to bear the cost of treatment), (ii) patient preference (proximity to their hometown or other personal reasons), (iii) hospital factors (lack of bed availability), and (iv) poor prognosis (deemed non-salvageable and requiring palliative care or needing long-term rehabilitation with a poor functional outcome). All the participants were counselled regarding the cost of expenses by our treating physician and the medical social worker, and only those patients choosing to leave DAMA from the Emergency Department (ED) were included. Initial assessment and management for all patients were as per the Advanced Trauma Life Support (ATLS) guidelines. All DAMA patients received the necessary life-saving interventions if indicated before leaving the hospital. These patients were followed up telephonically six months after their DAMA discharges as part of the trauma registry follow up to assess survival. Their socio-economic status was also calculated using the Modified Kuppuswamy Scale⁹.

All the continuous variables are reported using mean (SD) or median (IQR). All categorical variables are reported using frequency and percentages. Pearson chi-square test was used to find the association between categorical variables. Independent sample t-test or Mann-Whitney U test was used to compare the continuous variables between DAMA and non-DAMA groups. Step-wise univariable and multivariable logistic regression was done to assess the predictors of DAMA. Odds ratios (OR) with 95 per cent confidence interval (CI) are reported. A P<0.05 was considered statistically significant. All analysis was performed using STATA version 16.0 (StataCorp LLC, College Station, Texas, USA).

Results

A total of 2486 trauma patients were enrolled on the trauma registry during the study period, out of whom the number of DAMA patients was 174 (6.9%). The DAMA (n=174) and non-DAMA (n=2312) cohorts were comparable in terms of age and sex (Table I). Blunt injury was much more common in both groups than penetrating (99.5% DAMA vs. 97.5% non-DAMA). The most common mechanism of injury was RTI (73.9% DAMA patients and 72.8% non-DAMA patients). The majority of the patients were received as referred cases, having received primary treatment or care at another centre. Both groups were comparable in terms of baseline vital signs. The DAMA patients had a lower mean Glasgow Coma Score (GCS) (11.8±5.0 DAMA patients and 14.3±2.4 non-DAMA patients, P value <0.001) and a higher median ISS (9, IQR 4-13 of DAMA patients and 4, IQR 3-9 of non-DAMA patients, P value 0.001).

Descriptive analysis revealed that patients with a predominant Traumatic Brain Injury (TBI) accounted for 43.1 per cent of all DAMA patients. Among the DAMA patients, the most common reason cited was financial constraints (42.5%). The SES evaluation of these patients showed that most were in the upper-lower class (52.8%). Telephonic follow up could be completed for 159 patients (91.4%). It revealed that 57 patients did not survive (35.8%). Of 123 patients (77.4%) who continued treatment at another centre, 67 (54.5%) did so at a government centre. Under half of the DAMA patients who could be followed up (46.3%) required surgical intervention at another centre (Table II).

Univariable analysis revealed that the following parameters were found to increase the odds of DAMA: respiratory rate <18 breaths/min (OR 2.38, 95% CI 1.3–4.29), heart rate >95 beats/min (OR 1.73, 95% CI 1.26–2.37), ISS >8 (OR 1.71, 95% CI 1.24–2.35), GCS <15 (OR 3.47, 95% CI 2.41–4.98), having a predominant TBI (OR 2.6, 95% CI 1.9–3.6) and having predominant vertebral fractures/SCI (OR 2.5, 95% CI 1.4–4.5). Step-wise multivariable logistic regression was done for these positive parameters. Three variables were found to increase the odds of DAMA-GCS <15 (OR 2.6, 95% CI 1.4–4.8), predominant TBI (OR 1.99, 95% CI 1.33–2.98), and predominant vertebral fractures/SCI (OR 2.82, 95% CI 1.46–5.41) (Table III).

Discussion

The percentage of DAMA patients from our centre was significantly higher than in four similar retrospective studies from trauma centres in a high-income country, where the frequency varied from 0.76 to 1.8 per cent^10–13. The frequency, however, was lower when compared to general DAMA data from other hospitals within the country^14,15. The majority of our participants were middle-aged men, and blunt injuries were more common, and RTI was the most common mechanism of injury. All these findings were consistent with previously published institutional data¹⁶. Ours is a referral centre; only a few patients were received as primary cases. This usually means patients were taken to a primary or secondary care hospital before reaching our centre. With 87.9 per cent of DAMA patients being referred, this would mean that most of these patients received care in at least three hospitals, which is alarming. The DAMA patients also had a higher median ISS score, suggesting that they were more severely injured. Although descriptive data indicate that the majority of patients with polytrauma with high ISS managed to continue treatment at our centre, those with injuries requiring long-term care, such as patients with TBI or SCI, tended to leave AMA, likely because of the reasons enumerated below.

Descriptive analysis of DAMA patients revealed that the predominant reason for DAMA was financial, and most were in the lower or upper-lower socio-economic strata. This is similar to the data from a low-income country¹⁷. TBI was the predominant injury among DAMA patients. Logistic regression revealed that having a GCS < 15 for any reason and having a predominant TBI or vertebral fractures/SCI were the main predictors of DAMA. This is similar to three studies from high-income countries where DAMA was higher among patients with TBI, with a range of 1.8 to 2.8 per cent^15–17. One study from a government trauma centre in India has estimated that the median cost of in-hospital care for patients with a predominant head injury is INR 122,666 (USD 13,571), with 96 per cent of this cost being subsidized by the government¹⁸. However, barring a few States, currently, there are no centralized schemes available for trauma patients opting for care in private institutions. Private healthcare insurance coverage is also poor in India, reaching only 29 per cent of households^18,19. Even among patients with insurance, many of them do not cover the spectrum of care required for these patients²⁰. Patients with TBI and spinal cord injuries usually have an extended hospital course and need long-term rehabilitation, which many patients are unable to afford^21,22. These patients then opt for DAMA and seek affordable care in government or other hospitals. In addition to the patients themselves, there is a sizeable financial burden secondary to loss of income and catastrophic spending during the treatment period. Many of these patients are investigated repeatedly at each centre, and there is no continuity of care. This is in addition to the cost of each transfer, which usually comes as an out-of-pocket expenditure.

Follow up of DAMA patients revealed that nearly one-third did not survive. Most of the patients who survived completed their care at a government centre. A good number of patients also went toother private hospitals, likely because of two reasons, one is the coverage of specific trauma procedures under government schemes in the neighbouring State, and the other includes patients injured while travelling on the national highway, opting to shift to a centre closer to their hometown.

Private hospitals may be the most accessible to many patients immediately post-trauma. Many States within the country have started schemes to assist in immediate care for trauma victims in private hospitals²³. While this is a welcome step, the results of our study indicate that the benefits do not extend to patients who require long-term hospitalization and rehabilitation. We propose that Government subsidization, a public-private partnership economic model, and universal mandatory injury insurance coverage for vehicle owners could be a possible solutions to reduce health inequity and the DAMA rate. To ensure the standard of care, all trauma centres across the country, both public and private, should adhere to a standardized verification process, and only such centres should be eligible for governmental support and insurance coverage²⁴. Despite an extensive data search, the authors were unable to find a legal definition of terminologies like DAMA in the country, although they are used quite frequently. Standardizing these terms would be a welcome step to ensure that these patients are adequately captured in the system and their problems are highlighted.

Our study had several limitations. Being a single-institution study, the population that we cater to may not be generalizable to the entire country. Being a retrospective study, many variables are missing and need further elaboration in future studies, including - the socioeconomic status of both groups, identifying the cause of death for DAMA patients to know if it was preventable, the variability in the length of stay at our hospital before DAMA, ASIA grading of patients with SCIs, and the actual out-of-pocket expenses for patients as a result of leaving DAMA. Another limitation is that being a not-for-profit charitable hospital, many patients with financial difficulties still received complete treatment due to institutional support. Hence, the DAMA rates may have been higher and do not accurately represent the affected population. DAMA has not been previously studied among trauma patients in our country, and our study could sensitize the need for more extensive data to be collected nationwide to understand the problem better and suggest possible solutions.

Overall, our study findings highlight three important aspects of trauma care. One is that poor patients are often unable to afford comprehensive trauma care at private trauma centres. Secondly, the majority of these patients have TBI or SCI. Third, DAMA patients are understudied and usually excluded from data and clinical audits but appear to have poor outcomes. Future studies should aim to understand the cost burden and alleviate the causes. Policymakers should focus on achieving health equity to reduce the rate of DAMA and improve overall patient outcomes.