Identifying Age-Specific Risk Factors for Poor Outcomes After Trauma With Machine Learning.

INTRODUCTION: Risk stratification for poor outcomes is not currently age-specific. Risk stratification of older patients based on observational cohorts primarily composed of young patients may result in suboptimal clinical care and inaccurate quality benchmarking. We assessed two hypotheses. First, we hypothesized that risk factors for poor outcomes after trauma are age-dependent and, second, that the relative importance of various risk factors are also age-dependent. METHODS: A cohort study of severely injured adult trauma patients admitted to the intensive care unit 2014-2018 was performed using trauma registry data. Random forest algorithms predicting poor outcomes (death or complication) were built and validated using three cohorts: (1) patients of all ages, (2) younger patients, and (3) older patients. Older patients were defined as aged 55 y or more to maintain consistency with prior trauma literature. Complications assessed included acute renal failure, acute respiratory distress syndrome, cardiac arrest, unplanned intubation, unplanned intensive care unit admission, and unplanned return to the operating room, as defined by the trauma quality improvement program. Mean decrease in model accuracy (MDA), if each variable was removed and scaled to a Z-score, was calculated. MDA change ≥4 standard deviations between age cohorts was considered significant. RESULTS: Of 5489 patients, 25% were older. Poor outcomes occurred in 12% of younger and 33% of older patients. Head injury was the most important predictor of poor outcome in all cohorts. In the full cohort, age was the most important predictor of poor outcomes after head injury. Within age cohorts, the most important predictors of poor outcomes, after head injury, were surgery requirement in younger patients and arrival Glasgow Coma Scale in older patients. Compared to younger patients, head injury and arrival Glasgow Coma Scale had the greatest increase in importance for older patients, while systolic blood pressure had the greatest decrease in importance. CONCLUSIONS: Supervised machine learning identified differences in risk factors and their relative associations with poor outcomes based on age. Age-specific models may improve hospital benchmarking and identify quality improvement targets for older trauma patients.

Patient-Reported Outcomes Following Severe Abdominal Trauma: A Secondary Analysis of the Damage Control Laparotomy Trial.

INTRODUCTION: Little is known about patient-reported outcomes (PROs) following abdominal trauma. We hypothesized that patients undergoing definitive laparotomy (DEF) would have better PROs compared to those treated with damage control laparotomy (DCL). METHODS: The DCL Trial randomized DEF versus DCL in abdominal trauma. PROs were measured using the European Quality of Life-5 Dimensions-5 Levels (EQ-5D) questionnaire at discharge and six months postdischarge (1 = perfect health, 0 = death, and <0 = worse than death) and Posttraumatic Stress Disorder (PTSD) Checklist-Civilian. Unadjusted Bayesian analysis with a neutral prior was used to assess the posterior probability of achieving minimal clinically important difference. RESULTS: Of 39 randomized patients (21 DEF versus 18 DCL), 8 patients died (7 DEF versus 1 DCL). Of those who survived, 28 completed the EQ-5D at discharge (12 DEF versus 16 DCL) and 25 at 6 mo (12 DEF versus 13 DCL). Most patients were male (79%) with a median age of 30 (interquartile range (IQR) 21-42), suffered blunt injury (56%), and were severely injured (median injury severity score 33, IQR 21 - 42). Median EQ-5D value at discharge was 0.20 (IQR 0.06 - 0.52) DEF versus 0.31 (IQR -0.03 - 0.43) DCL, and at six months 0.51 (IQR 0.30 - 0.74) DEF versus 0.50 (IQR 0.28 - 0.84) DCL. The posterior probability of minimal clinically important difference DEF versus DCL at discharge and six months was 16% and 23%, respectively. CONCLUSIONS: Functional deficits for trauma patients persist beyond the acute setting regardless of laparotomy status. These deficits warrant longitudinal studies to better inform patients on recovery expectations.

Ketamine For Acute Pain After Trauma (KAPT): A Pragmatic, Randomized Clinical Trial.

INTRODUCTION: Non-narcotic intravenous medications may be a beneficial adjunct to oral multimodal pain regimens (MMPRs) which reduce but do not eliminate opioid exposure and prescribing after trauma. We hypothesized that the addition of a sub-dissociative ketamine infusion (KI) to a standardized oral MMPR reduces inpatient opioid exposure. METHODS: Eligible adult trauma patients admitted to the intermediate or intensive care unit were randomized upon admission to our institutional MMPR per usual care (UC) or UC plus sub-dissociative KI for 24 to 72 hours after arrival. The primary outcome was morphine milligram equivalents per day (MME/d) and secondary outcomes included total MME, discharge with an opioid prescription (OP%), and rates of ketamine side effects. Bayesian posterior probabilities (pp) were calculated using neutral priors. RESULTS: A total of 300 patients were included in the final analysis with 144 randomized to KI and 156 to UC. Baseline characteristics were similar between groups. The injury severity scores for KI were 19 [14, 29] versus UC 22 [14, 29]. The KI group had a lower rate of long-bone fracture (37% versus 49%) and laparotomy (16% versus 24%). Patients receiving KI had an absolute reduction of 7 MME/day, 96 total MME, and 5% in OP%. Additionally, KI had a relative risk (RR) reduction of 19% in MME/day (RR 0.81 [0.69 - 0.95], pp = 99%), 20% in total MME (RR 0.80 [0.64, 0.99], pp = 98%), and 8% in OP% (RR 0.92 [0.76, 1.11], pp = 81%). The KI group had a higher rate of delirium (11% versus 6%); however, rates of other side effects such as arrythmias and unplanned intubations were similar between groups. CONCLUSION: Addition of a sub-dissociative ketamine infusion to an oral MMPR resulted in a decrease in opioid exposure in severely injured patients. Sub-dissociative ketamine infusions can be used as a safe adjunct to decrease opioid exposure in monitored settings. LEVEL OF EVIDENCE: I; Therapeutic/Care Management.

Impact of Intra-Operative Shock and Resuscitation on Surgical Site Infections After Trauma Laparotomy.

Background: Patients undergoing trauma laparotomy experience high rates of surgical site infection (SSI). Although intra-operative shock is a likely contributor to SSI risk, little is known about the relation between shock, intra-operative restoration of physiologic normalcy, and SSI development. Patients and Methods: A retrospective review of trauma patients who underwent emergent definitive laparotomy was performed. Using shock index and base excess at the beginning and end of laparotomy, patients were classified as normal, persistent shock, resuscitated, or new shock. Univariable and multivariable analyses were performed to identify predictors of organ/space SSI, superficial/deep SSI, and any SSI. Results: Of 1,191 included patients, 600 (50%) were categorized as no shock, 248 (21%) as resuscitated, 109 (9%) as new shock, and 236 (20%) as persistent shock, with incidence of any SSI as 51 (9%), 28 (11%), 26 (24%), and 32 (14%), respectively. These rates were similar in organ/space and superficial/deep SSIs. On multivariable analysis, resuscitated, new shock, and persistent shock were associated with increased odds of organ/space SSI (odds ratio [OR], 2.2; 95% confidence interval [CI], 1.3-3.5; p < 0.001) and any SSI (OR, 2.0; 95% CI, 1.4-3.2; p < 0.001), but no increased risk of superficial/deep SSI (OR, 1.4; 95% CI, 0.8-2.6; p = 0.331). Conclusions: Although the trajectory of physiologic status influenced SSI, the presence of shock at any time during trauma laparotomy, regardless of restoration of physiologic normalcy, was associated with increased odds of SSI. Further investigation is warranted to determine the relation between peri-operative shock and SSI in trauma patients.

Opioid Risk Tool, in-hospital opioid exposure, and opioid demand predict pain outcomes following traumatic injury.

Prescribed opioids are a mainstay pain treatment after traumatic injury, but a subgroup of patients may be at risk for continued opioid use. We evaluated the predictive utility of a traditional screening tool, the Opioid Risk Tool (ORT), and two other measures: average in-hospital milligram morphine equivalents (MME) per day and an assessment of opioid demand in predicting pain outcomes. Assessments of pain-related outcomes (pain intensity, interference, injury-related stress, and need for additional pain treatment) were administered at 2 weeks and 12 months post-discharge in a sample of 34 patients hospitalized for traumatic injury. Bayesian linear models were used to evaluate changes in responses over time as a function of predictors. High-risk ORT, higher MME per day, and greater opioid demand predicted less change in outcomes over time. This report provides first evidence that malleable factors of opioid and opioid demand have utility in predicting pain outcomes following traumatic injury.