Abstract
Background
Research has shown improvements in patient care and outcomes with addition of a rounding geriatrician. The purpose of this study was to determine if addition of a hospitalist consultation improved patient outcomes.
Methods
A retrospective review was conducted of all trauma patients, ≥65 years, before (n=481) and after (n=430) addition of a hospitalist consultant. Data included were demographics, comorbidities, injury severity, blood pressure, laboratory levels, pain control methods, ICU and ventilator requirements, complications, hospital length of stay, mortality, preexisting wishes, and 30-day readmission.
Results
Adding a hospitalist consultation did not improve blood glucose or blood pressure control. It decreased narcotics-only use (36.0% vs 73.8%) while increasing multimodal pain control use (51.8% vs 14.8%, P<.001) and testing of HbA1c (7% vs .6%, P<.001). There was also increased knowledge of patient resuscitation status preferences (29.1% vs 12.9%, P<.001).
Conclusions
This article does not support use of routine hospitalist consultation in the geriatric trauma population. However, with study limitations, we continue to evaluate hospitalist utility and will adjust our daily rounds to more closely match prior studies.
Key Takeaways
• Outcomes in patients ≥65 were evaluated before and after addition of a hospitalist consultation. • While blood pressure and glucose control were not improved by hospitalist consultation, hospitalist consultation decreased narcotics-only use (36.0% vs 73.8%) and increased multimodal pain control use (51.8 vs 14.8%, P < .001). • Addition of a hospitalist consultation to geriatric trauma patients was not supported by these study results, but with study limitations including implementation issues and communication issues, we are continuing the practice with plans to reevaluate the efficacy in a more stable team.
Introduction
The proportion of the United States population over the age of 65 is expected to increase from 15% to 24% from 2014 to 2060. 1 As unintentional injury is the seventh leading cause of death for those over age 65, it will be necessary for trauma centers to be ready to treat and manage increased numbers of geriatric patients. 2 In response to the Ad Hoc Geriatric Committee, Kozar et al 3 identified the need for an interdisciplinary approach to the care of the geriatric trauma patient. Brooks and Peetz 4 described the physiologic differences that make caring for geriatric patients different from caring for their younger counterparts, as well as to discuss preexisting conditions that make this group more vulnerable to trauma and less able to tolerate the stress of trauma. Due to the more vulnerable nature of geriatric patients, some have advocated for inclusion of a geriatrician in the care of elderly trauma patients. Prior studies that initiated geriatric trauma services have shown improvements in time to definitive operation, intensive care unit (ICU) length of stay (LOS), hospital LOS, urinary tract infections (UTIs), delirium rates, and mortality.5-7
Therefore, in early 2017, the multidisciplinary trauma team at our tertiary care, level 1 trauma center implemented a change for all geriatric trauma patients: any patient 65 years of age or older would have an automatic consult to a hospitalist group with experience in caring for geriatric patients. Hospitalists were utilized for the geriatric consultations as dedicated geriatricians were not available at this institution. Prior to 2017, the trauma team, consisting of trauma surgeons, general surgery residents, a trauma nurse practitioner, and case managers, oversaw the management of trauma patients and consulted other medical services on an individualized basis. After the implementation of the automatic consultation, the trauma team continued to lead the team taking care of the patients with input from the hospitalist group regarding management of the geriatric patients’ comorbidities. The goal of this study was to evaluate changes in the care and outcomes of geriatric trauma patients before and after the implementation of a hospitalist consult.
Methods
This study was approved by the Ascension Via Christi Hospitals Wichita, Inc. Institutional Review Board. A 2-year retrospective chart review was conducted of all geriatric trauma patients, aged 65 years or older, who presented to our American College of Surgeons Committee on Trauma verified level 1 trauma center. The patients from January 2016 to December 2016 were used as the control group of geriatric trauma patients without a mandated hospitalist consultation. The patients from April 2017 to December 2017 were used as the comparison group of geriatric trauma patients with an automatic hospitalist consultation. The hospitalists were internal medicine physicians employed by the hospital. While they are not fellowship-trained geriatricians, the majority of their usual patient population is over 65 years of age and will hereafter be referred to as hospitalists. We did not include patients that presented in the first 3 months after the initiation of the protocol to account for adjustments in workflow that may bias the comparison. Patients that did not meet inclusion criteria (patient aged less than 65, or medical record not available) or that were admitted to the hospital by other services were also excluded.
The implementation of the G65 concept was very challenging from the beginning and may benefit from a description in some greater detail. The first challenge was the lack of qualified true gerontologists, which was addressed by finding hospitalists who were interested in geriatric care. Using this model led to additional difficulties in that the daily flow would not allow the actual hospitalists to round physically with the trauma team. Additionally, they are based on a day/night schedule with the individual changing every 3 days. These changes led to more variability as the care was physician specific and could change after a 3-day period and was less standardized as there was no agreement of standardized blood pressure or blood glucose protocols.
Patient data was obtained from the trauma registry database as well as from patient medical records available in electronic and paper format. The information collected from each patient included patient demographics [age, gender, and race]; trauma level; comorbidities; injury severity score (ISS); and abbreviated injury severity (AIS) scores on admission; Glasgow Coma Scale (GCS) score on admission; vital signs on admission and throughout hospital stay (blood pressure and heart rate); ICU admission and LOS; need for mechanical ventilation and ventilator days; complications; hospital LOS; mortality; preexisting wishes regarding end of life care; in-hospital procedures and interventions; glucose levels; creatinine and BUN levels; urinalysis results; lactic acid; HbA1c; pain control methods (such as IV/PO narcotic medications, NSAID usage, and analgesic patch usage); 30-day readmission; and mortality. Patients were evaluated for control of blood glucose and blood pressure during hospitalization. We defined hyperglycemia as a blood glucose of greater than 200 mg/dL. Hypertension was defined as a systolic blood pressure of greater than 150 mmHg or a diastolic blood pressure of greater than 90 mmHg per the Joint National Committee 8 guidelines. There were no management guidelines for glucose control or blood pressure control in either period. Glucose and blood pressure control were at the discretion of the treating physicians.
Statistical Analysis
Data was compiled, evaluated, and summarized by calculating means and standard deviations for normally distributed continuous data, medians and interquartile ranges for non-normally distributed continuous data, and proportions for categorical data. An Independent Samples T-test was used to compare normally distributed continuous data. Skewed continuous data was compared with the Mann-Whitney U-Test. Chi-square analysis was used to compare categorical data. All analyses were run as two-tailed tests and results of analyses are considered significant if the resultant P value is less than or equal to .05. Analyses were run using SPSS release 19.0 (IBM Corp., Armonk, NY).
Results
Demographics and Comorbidities of Trauma Patients Treated and Evaluated Without (Period I) or With a Hospitalist Consultation (Period II) a .
aData are presented as the number (%) or median (IQR).
Injury Severity and Admission Vital Signs of Trauma Patients Treated and Evaluated Without (Period I) or With a Hospitalist Consultation (Period II) a .
aData are presented as the number (%), median (IQR), or mean ± standard deviation.
Laboratory Data, Hospital Outcomes, Discharge Destination, and 30-Day Readmission Rates of Trauma Patients Treated and Evaluated Without (Period I) or With a Hospitalist Consultation (Period II) a .
aData are presented as the number (%), median (IQR), or mean ± standard deviation.
Other laboratory values including creatinine, lactic acid, and blood urea nitrogen were similar between study periods (Table 3). However, there was a statistically significant difference in rates of testing of HbA1c from period I to II, with 3 patients (.6%) being tested from period I and 30 (7.0%) from period II (P < .001). Similarly, there were lower rates of urine analyses in period I (37.6% vs 52.8%; P < .001). However, there was not a statistically significant difference in rates of UTI between period I and II for those undergoing urinalysis (23.8% vs 20.7%, respectively; P = .460). While more patients were admitted to the ICU in period I (68.4% vs 57.9%; P < .001), ICU length of stay was not significantly different for those patients admitted to the ICU. More patients were mechanically ventilated in period I compared to period II (10.2% vs 6.5%; P = .046), but days on the ventilator were not significantly different. Hospital LOS, overall mortality, discharge disposition, and readmission rates were similar between groups, but there were more patients who were made comfort care in period I (8.3% vs 4.9%; P = .039).
Operative Intervention and Pain Control Medication Usage of Trauma Patients Treated and Evaluated Without (Period I) or With a Hospitalist Consultation (Period II).
In period I, 5% of patients received no pain medications compared to only 44.7% of patients in period II (P < .001, Table 4). Forty-nine percent in period I received pain medications, compared to 41% in period II. Of those who used pain medication, 73.8% in period I used only narcotics, while 11.4% used non-narcotics only and 14.8% used narcotics plus one or two non-narcotic adjuncts. Of those who used pain medication in period II, 36% used narcotics exclusively, while 12.4% used non-narcotics only and 51.6% used narcotics plus one or more non-narcotic adjunct.
Preexisting Wishes of Trauma Patients Treated and Evaluated Without (Period I) or With a Hospitalist Consultation (Period II).
Discussion
In our study, we did not see the improvement in major outcomes of elderly trauma victims with addition of hospitalists that were observed in previous studies. We did not find a significant decrease in mortality, 30-day readmission rates, or change in discharge disposition. In our study, we observed a 2.1%, nonsignificant decrease in mortality with addition of a hospitalist consultation (P = .231). Our findings were similar to those of Mangram et al, 5 Olufajo et al, 8 and Lenartowicz et al 6 who also observed nonsignificant decreases in mortality of 1.9%, 4.1%, and 2.3%, respectively, after addition of a mandatory geriatrician consult to elderly trauma patients. Also, as in our study, Olufajo et al did not observe an improvement in the 30-day readmission rate with addition of a geriatrician consult (16.9% vs 14.9%, P = .6). In contrast, Bradburn et al 7 instituted a high-risk geriatric protocol in geriatric trauma patients who had a high-risk injury and at least one comorbidity. Their geriatric protocol included the addition of a mandatory consult to a geriatrician. In those patients in the geriatric protocol, they found a reduction in mortality in the patients who had the geriatric protocol after adjusting for other variables, with an odds ratio of .63 [.39-.99] (P = .046).
Additionally, when evaluating ICU LOS and hospital LOS, we did not find a significant improvement by addition of a hospitalist consult. In this study, ICU LOS was 2 days for both study periods and hospital LOS increased by 1 day from period I to period II (3 vs 4 days, respectively). In Mangram’s study, they observed a decrease in ICU LOS by 2 days by addition of a geriatrician consultation (5.2 vs 3 days) and a similar decrease in hospital LOS (7 vs 4.8 days). 5 In Olufajo’s study, the hospital length of stay had a nonsignificant decrease from 6.4 to 5.9 days before and after the mandatory geriatrician consult. 8 In Lenartowicz’s study, hospital LOS was evaluated and found to be 19.4 days for the pre-GTCS group and 15.4 days in the post-GTCS group (P = .13). 6 It should be noted again that the patients in that study had higher average ISS and LOS than the patients in our study, making their findings difficult to compare to ours. With a short length of stay in the ICU and hospital for our first study period, it is not surprising to the authors that there was not a significant decrease for either LOS by addition of a hospitalist consultation.
Our study also did not demonstrate a change in the discharge disposition of our patients before and after the automatic consultation to a hospitalist. Nearly half of our patients in both groups went on to a skilled nursing equivalent upon discharge. This is similar to Mangram’s study where they observed no difference between their 2 study populations and noted more than 50% of patients went to an interim facility for rehabilitation or continued medical care. 5 Lenartowicz et al 6 compared those patients who came in from home in their pre- and post-GTCS groups found a significant decrease in the proportion of patients that went to long-term care facilities upon discharge in the post-GTCS group (6.5% vs 1.7%; P = .03).
Unique to this study, we attempted to determine times to blood glucose control and blood pressure control. We were unable to find a significant difference between the 2 groups with regard to control of blood glucose. Among those in the inconsistent group, we evaluated number of days to achieve blood glucose control: there was no significant difference between the 2 groups’ time to achieve blood glucose control. Similarly, we did not observe a significant difference between the groups regarding blood pressure control. Within the inconsistent control of the blood pressure group, we again looked at days to achieve blood pressure control but did not observe a significant difference between the 2 groups. Of note, although all patients in period II were seen by the hospitalists, there was no protocol regarding blood glucose control or blood pressure control. This likely contributed to the limited impact that the hospitalist team had on these measures.
While the rates of operative intervention, ICU admission, and mechanical ventilation were statistically different between period I and II, the addition of the automatic hospitalist consultation is likely not the underlying cause. The hospitalists have most input on patients when they are out of the ICU as the trauma intensive care team takes full responsibility of the patients in the ICU.
We did find improvements in the usage of multimodal pain control and decreased narcotic pain medication usage during the time of this study. In this case, we did have protocols in place due to a hospital-wide effort to curb the use of narcotics and encourage multimodal pain control. It is hard to determine how much the hospital-wide effort affected this as compared to the hospitalist consult.
With the help of the hospitalist consultation, we were able to define patient’s resuscitation status more clearly in period II than I. This could be due to the enhanced physician time to discuss options leading to greater patient comfort and confidence in determining resuscitation status rather than the limited time and direct questions asked by the trauma surgeons and surgical residents. Similar to our findings, Olufajo et al 8 found that the rates of DNR/DNI code status increased after the initiation of the mandatory geriatrician consult (10.2% vs 38.2%, P<.01).
Limitations
This was a retrospective study at 1 institution, thereby limiting generalizability. There were differences between the 2 groups at baseline; for example, there were more patients with congestive heart failure in period II and more patients underwent operative intervention in period II. The hospitalists did not round with the trauma team and did not attend multidisciplinary rounds routinely, and thus, communication issues may have contributed. There was not a set geriatric floor or unit unlike in other studies. There are no protocols for hypertension and diabetes and thus were provider specific. Although the Internal Medicine physicians who were consulted do frequently see geriatric patients, they did not have advanced training or fellowships in geriatrics.
Conclusions
On review of our data, we found no support in positive outcomes with regard to mortality, length of stay, discharge disposition, or 30-day readmission rate from addition of the hospitalist service. Unique to this study, we did not see decreased time to blood glucose or blood pressure control. We did observe positive outcomes in the form of decreased ICU admission and mechanical ventilation, although the etiology of these is not clear to the authors. This article does not support use of routine hospitalist consultation in the geriatric trauma population.
Footnotes
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.