Hospitalist Research Presentations

1. Hospitalist Research Presentations September 30, 2011

2. Goals • Increase awareness (among hospitalists, residents, students, other IM faculty) about research that Hospitalist Division is doing. • Encourage other hospitalist to get involved because they will see how exciting and fun this is. • Discuss obstacles and challenges that we can all learn from. • Help investigator by having questions/critiques

3. Methodolgy • 4 presentations • Dr. DeeptiRao – Teaching physical diagnosis to medical students • Dr. Kendall Rogers – Quality Improvement -Glycemic Control • Dr. Lenny Noronha - Observed clinical encounters of medical students • Dr. Rush Pierce – Fall-related injury in hospital • Each presentation 8 mins, 2 mins questions

4. Student vs moderator driven physical exam rounds

5. Investigators • Deepti Rao, MD • Rishi Menon, MD • Rush Pierce, MD

6. Background • Students in their post clerkship evaluations requested physical exam “rounds” • Dr. Menon voiced interest, independently, in performing pe rounds with students • Dr. Comerci had, in the past, trialed a physical exam rounds method where students directed the rounds • A literature search had no evidence regarding the method for teaching “physical exam rounds”

7. Research Aims • To find out if students find value in physical exam rounds • To find out if students find value in student versus moderator driven physical exam rounds • To find out if students become proficient in certain aspects of the physical exam • To find if physical exam rounds might have an effect on test scores

8. Methods • Weekly physical exam rounds either student or moderator driven at university hospital • First week batcave • Second week cardiovasc • Third week pulm • Fourth week gi and posttests

9. Methods • Usually one hour long • Text: McGee Evidence Based Physical Exam • Whoever in charge leads others through exam, case and discussion • In student driven the students divide up weeks

10. Example

11. Methods--pretest • Rate your confidence with the below physical exam maneuvers on a scale from 1-5 (5 being very confident and 1 being not confident at all). • Assessment of JVD • Assessment of ascites • Percussion for fluid level • Assessment for fluid wave • Percussion of the lung

12. Methods--posttest • Please assess the following on a scale from 1-5 with 5 being the best. • I found this activity educationally valuable. • I improved my physical exam skills through this course. • I found the scheduling/timing of this course fit into my day. • The moderator(s) was effective in finding and demonstrating • physical exam findings. • Rate your confidence with the below physical exam maneuvers. • Assessment of JVD • Assessment of ascites • Percussion for fluid level • Assessment for fluid wave • Percussion of the lung • How many of these sessions did you attend? • How many patients with these diagnoses did you see on physical exam rounds? • Congestive heart failure • Heart murmur • Pneumonia • Pleural effusion • Liver cirrhosis with ascites

13. Methods—PE posttest • JVD • Head at 30 degrees or higher • Assessment made from patient’s right • Head turned to left • Tangential light • Accurate assessment of marked JVD • Ascites (may choose 1 of 2 methods of assessment) • Shifting dullness • Percussion of both flanks • Patient turned slightly • Line of dullness reassessed from up and down • Fluid wave • Hand placed midline • Bilateral had placement to assess fluid wave • Percussion of lung • Side to side technique • Did not percuss over scapula • At least 3 pairs

14. Timeline • 07/2010 – initial meetings between Rishi and Rush and I • 09/2010 – IRB submitted (Exempt study) • 10/2010—IRB approval • 01/2011 – study ongoing • 01/2011—study finished and data will be reviewed

15. Problems/limitations • Difficulty in finding time • Some students not interested • Some weeks very hard to find interesting findings

16. Quality of Care Initiative: Blood glucose timing, food delivery and insulin administration

17. Investigators • Kendall M. Rogers MD CPE FACP SFHM, Primary Investigator, Department of Internal Medicine • Rosa Matonti RN BSN CDE, Co-Investigator, Glycemic Control Clinical Nurse Specialist, University of New Mexico HSC • Susan Hoeppner RN BSN, Co-Investigator, SAC Nurse, University of New Mexico HSC

18. Background • Hyperglycemia is associated with poor outcomes in a broad range of hospitalized patients, and several studies demonstrate improved outcomes with improved glycemic control. • Hospitalization presents a frequently missed opportunity to diagnose diabetes, identify those at risk for diabetes, and to optimize the care of patients with diabetes via education and medical therapy. • Despite authoritative guidelines and effective methods to achieve good glycemic control safely, poor glycemic control, suboptimal medication regimens, incomplete patient education, and uneven communication with outpatient care providers are prevalent problems in medical centers. • Inpatient insulin timing with meal delivery and other nursing processes are a core, but often overlooked, component to achieving glycemic control

19. Background • On early evaluations, BG monitoring did not coincide with meal delivery • Insulin administration was not coordinated with meal delivery and/or BG monitoring and often would occur greater than 30 minutes after the delivery of the meal tray • There were no protocols in place for standards of practice regarding BG timing and insulin administration • We believed these process discrepancies were resulting in worse BG outcomes

20. Research Aims • Evaluate the timing of our current in-patient process between BG monitoring, insulin administration and bedside meal delivery • Create workflows that would allow for this process to be completed within a 30 minute window of meal time delivery • Create protocols to guide staff with the most current evidence based practice • Measure the outcome differences in patients before and after this intervention

21. Methods • Controlled before and after study of a 11 month PDSA Cycle Quality Improvement Project • Inclusion: all patients receiving insulin on a adult inpatient unit where approximately one-third of patients receive insulin • Time in motion studies and process mapping • RALS glucose database

22. Methods • Literature review • Development and approval of policies • Time in motion studies to evaluate current processes • Process mapping by front line workers • Nurse and front line staff driven development of processes • Multiple interventions • Multiple cycles with time in motion studies

23. Interventions • Coordination of communication with cafeteria • Development and dissemination of written policy with expectations of insulin admin • 1st phase: improved tech/nurse coordination for testing and insulin delivery • 2nd phase: Nurse driven process of CBG check, meal delivery, and insulin administration

24. Results

25. Baseline Results • Staff obtained BGs >30 minutes prior to meal 49% of the time, ranging from 166 minutes before meal to 98 minutes after meal • Meal timing was inconsistent; delivery to floor varied by 15-30 minutes each day, followed by tray audits taking ≥15 minutes on average

27. Mean Time to Insulin Admin from CBG check

30. Glucose Control Comparison

31. Limitations • The time in motion studies took place over 5 days and BG numbers were utilized as the n value • Only a sub acute unit was studied not generalizable to other units • Study completed in an Academic Level 1 Trauma medical center and results can not be generalized to other centers • Staff were aware of when the study was conducted • Study tools were rudimentary and not precise. Cell phones used for timing.

32. Conclusions • A nurse driven process of CBG check, meal delivery, and insulin administration achieved process goal of insulin admin <30 min of CBG check. • When process goal achieved, mean glycemic control improved for this patient population

33. Predicting Hospital Fall-Related Injury

34. Investigators • Rush Pierce, MD, (Investigator), Hospitalist • Michael Shirley, MD (Co-I), Resident • Emma Johnson (Co-I), MS1 • Dierdre Kearney, RN, MSN, nurse manager

35. Background • Inpatient falls are the source of a significant harm to patients and costs to hospitals. • 2 – 9% of adult patients who fall in the hospital suffer serious injury. • There is little evidence-based literature to guide physicians when assessing patients for fall-related injury.

36. Research Aims • Identify demographic and/or clinical factors which predict serious injury resulting from falls among hospitalized patients. • Understand what tests are used to assess injury after falls among hospitalized patients. • Judge the adequacy of documentation of the occurrence of falls and of evaluation of patients who have fallen in the hospital

37. Methods • Retrospective chart review • Inclusion: PSN reported fall in 2010 • Exclude: < 18 y/o, pregnant, prisoner • Chart review for fall related injury, demographic factors, clinical factors, tests ordered, documentation

38. Methods: 1. Identify demographic and/or clinical factors predicting injury • Injury vs. no injury • Demographic factors: Age, gender, hospital unit • Clinical factors: witnessed; hit head; LOC; change in MS; neck pain, HA, weakness/numbness, presence of Foley • Meds: (a) heparin, fondaparinux, Lovenox, warfarin, argatroban, IIb/IIIa inhibitors, or plavix; (b) narcotics, benzodiazepine, nifedipine, antihistamine • Lab: INR, platelet count

39. Methods: Injury and severity • (1) none: no significant discomfort and no increased LOS • (2) minor: minimal impact on patient care, may involve some clinical intervention, no lasting impact on patient outcome (eg, minor cuts, minor bleeding, swelling, pain, or minor contusions) • (3) moderate: moderate impact on patient care, some clinical intervention, and some lasting impact on the patient’s outcome (eg, excessive bleeding, lacerations requiring sutures, loss of consciousness) • (4) major: severe impact on patient care, major clinical intervention, and major impact (functional or cosmetic impairment) on patient’s outcome (eg, fractures, subdural hematomas, or cardiac arrest) • (5) death: death related or attributed to the fall.

40. Methods 2. What tests are used to assess injury after falls • Imaging studies obtained • Whether these studies indicated injury 3. Adequacy of documentation • Note specific to fall • Mentioned in same day/next day progress note • Mentioned in discharge summary

41. Timeline • 02/08/2011 – initial meeting w/resident • 03/11/2011 – IRB submitted (Exempt study) • 03/15/2011 – student recruited • 03/18/2011 - IRB approval • 04/20/2011 - student approved and completes training • 05/15/2011 – initial chart reviews done • 06/01/2011 – initial data to statistician • 06/01/2011 – request for data from J. Little (received in 2d) • 06/17/2011 – cleaned up data to statistician • 09/18/2011 – statistical results from statistician

42. Results • 298 PSN’s reporting falls in 2010 • Exclude pts < 18 y/o (2), pregnant pts (3), pts with no matching MR (7) • 286 falls, 251 pts • 152 males (61%), 99 females (39%) • 63 falls with injury (25%)

43. Falls by unit

44. Fall related to toileting

45. Falls with injury

46. Predictors of fall with injury

47. Documentation, all falls

48. Documentation, falls with injury

49. Problems/limitations • Methodologic Problems – hard to find Foley; few missing charts; two deaths close to fall; initial spreadsheet not clean; UHC database problems • Operational problems – did not anticipate delay for student to get approved; statistician on extended vacation/other projects • Limitations – observational study; PSN may not capture all falls; medication use for all pts different than patients who fall

50. Conclusions • One-half falls related to toileting • 25% of falls were associated with injury, 4% with mod/severe impact on patient care. • Over 60% of falls did not have any physician documentation; 30% of falls with injury