Glucometrics:    Assessing Quality in  Inpatient Glycemic Management

Glucometrics: Assessing Quality in Inpatient Glycemic Management PowerPoint PPT Presentation


  • 238 Views
  • Uploaded on
  • Presentation posted in: General

Download Presentation

Glucometrics: Assessing Quality in Inpatient Glycemic Management

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript


1. “Glucometrics:” Assessing Quality in Inpatient Glycemic Management

2. “Glucometrics”: Assessing Quality in Inpatient Glycemic Management Outline What is ‘quality’ inpatient glucose management? How should we measure it? Future directions

3. The 4 Spheres of a Quality Inpatient Glucose Management Program

4. A Hospital Priority Why? Enhance quality & patient safety Competitive advantage Cost savings The Joint Commission

7. Institution-Wide Educational Efforts

8. Patient Care Identification (& coding) of patients Policies & procedures Point-of-Care BG Testing Institutional BG Targets (ICU, Wards) Hypoglycemia protocol ICU IV insulin protocols Standardized SQ insulin order sets Patient education tools “Inpatient diabetes management team” Transition to outpatient care (access)

12. Example of a Glucometrics Report

13. Graphic Display of Glucometrics Data

16. Measuring Inpatient Glycemic Control: Special Issues Sample site (fingersticks, lab plasma glucose) Multiple measures during hypoglycemic or hyperglycemic “events” Varying time intervals of measurement Timing in relationship to meals Effects of IV fluids (dextrose) How to collect glucose measurements? How to analyze them? How to present data to clinicians/adminstrators?

17. The 4 Spheres of a Quality Inpatient Glucose Management Program Metrics Systematic review of hospital BG data Analytical models What’s the “HbA1c” for glucose control during a hospitalization?

18. Metrics Traditionally Used in the Inpatient Glucose Literature Raw blood glucose (BG) average % of BGs within a pre-specified range (80-110, 100-150, <180, <200 mg/dl) % of patients with a certain % of BGs within a pre-specified range Hypoglycemia rates (<40, <50, <60, <70 mg/dl) % of BGs % of patients Hyperglycemic excursions (>180, >200, >300 mg/dl) % of BGs % of patients

19. Glucometrics Project: Objectives Define inpatient glucose quality metrics Define ‘units of analysis’ Compare metric results for the different analytical units The objectives of our analysis were the following: Define useful BG quality metrics Define units of analysis Compare the results of BG measures when applied to the various units of analysis The objectives of our analysis were the following: Define useful BG quality metrics Define units of analysis Compare the results of BG measures when applied to the various units of analysis

20. Methods: Sample Yale-New Haven Hospital BG data BG data downloaded into relational database for analysis BG values Date / Time Patient ID Hospital ward Sample: One general medical ward’s March 2004 BG results (n=1,552) The analysis was performed using BG data from YNHH, a 944 bed university-affiliated tertiary regional referral center in New Haven, CT. BG results were down-loaded from bedside glucose meters into a relational database for analysis. Included in the glucose meter data along with each BG value (plasma-standardized) are the corresponding date and time of the result, the patient MRN (or patient “ID”) and the hospital ward. For convenience, we used a sample of one general medical ward’s BG results for one month (n=1,552 results). The analysis was performed using BG data from YNHH, a 944 bed university-affiliated tertiary regional referral center in New Haven, CT. BG results were down-loaded from bedside glucose meters into a relational database for analysis. Included in the glucose meter data along with each BG value (plasma-standardized) are the corresponding date and time of the result, the patient MRN (or patient “ID”) and the hospital ward. For convenience, we used a sample of one general medical ward’s BG results for one month (n=1,552 results).

21. Mean / Median BG % BG in “favorable” range (80 - 139 mg/dl) % Hyperglycemia (>300 mg/dl) % Hypoglycemia (<60 mg/dl) The quality performance metrics we tested were derived from a literature review of guidelines for favorable glucose control. We constructed these measures to allow comparisons between hospital wards as well as between individual patients. The measures are: Mean and Median BG % BG in favorable range (somewhat arbitrarily defined as 80-139 mg/dl) % Hyperglycemia (defined as >300 mg/dl) % Hypoglycemia (defined as <60 mg/dl) The quality performance metrics we tested were derived from a literature review of guidelines for favorable glucose control. We constructed these measures to allow comparisons between hospital wards as well as between individual patients. The measures are: Mean and Median BG % BG in favorable range (somewhat arbitrarily defined as 80-139 mg/dl) % Hyperglycemia (defined as >300 mg/dl) % Hypoglycemia (defined as <60 mg/dl)

22. “Ward” n = 1,552 “Patient Stay” n = 118 [13.2 BGs / stay] “Patient Day” n = 467 [3.3 BGs / day] To best understand the usefulness of our metrics as quality measures, we examined how they behaved in the context of different units of analysis. We defined these 3 units of analysis: The WARD, which looked at individual BG values in the sample, a total of 1,552. For instance the mean using this unit of analysis would be the average BG among all 1552 readings. The PATIENT, which was derived from grouping the individual values by each unique Patient ID. There were 118 patients in the sample with an average of 13.2 BG values per patient. This model highlights glycemic control for individual patients during hospitalization. For example, the mean using this unit of analysis would be the average of all the individual means for each patient. The PATIENT DAY was constructed by clustering BG values by Patient ID and by Hospital Day. There were a total of 467 days in which at least one of the 118 patients had at least one BG result, for an average of 3.3 values per day. Total days that BG were performed ranged from 1 to 31 days, with an average of 4.0 days. This model normalizes for the otherwise disproportionate contribution to the total BG values by patients with very long LOS. So, using the same example…the mean here would be the average of each patient-specific mean on each day in which BG was checked. So, once we agreed upon these analytical units, we then calculated the results of the metrics for all 3, represented by the 3 different denominators: N=1552 for the WARD level, N=467 for the PATIENT DAYs when BG were measured, and N=118 for the number of PATIENTS with BG. To best understand the usefulness of our metrics as quality measures, we examined how they behaved in the context of different units of analysis. We defined these 3 units of analysis: The WARD, which looked at individual BG values in the sample, a total of 1,552. For instance the mean using this unit of analysis would be the average BG among all 1552 readings. The PATIENT, which was derived from grouping the individual values by each unique Patient ID. There were 118 patients in the sample with an average of 13.2 BG values per patient. This model highlights glycemic control for individual patients during hospitalization. For example, the mean using this unit of analysis would be the average of all the individual means for each patient. The PATIENT DAY was constructed by clustering BG values by Patient ID and by Hospital Day. There were a total of 467 days in which at least one of the 118 patients had at least one BG result, for an average of 3.3 values per day. Total days that BG were performed ranged from 1 to 31 days, with an average of 4.0 days. This model normalizes for the otherwise disproportionate contribution to the total BG values by patients with very long LOS. So, using the same example…the mean here would be the average of each patient-specific mean on each day in which BG was checked. So, once we agreed upon these analytical units, we then calculated the results of the metrics for all 3, represented by the 3 different denominators: N=1552 for the WARD level, N=467 for the PATIENT DAYs when BG were measured, and N=118 for the number of PATIENTS with BG.

29. Here are the mean and median BGs for the 3 units of analysis. As you can see, there is little difference in the means. Also, the medians are smaller, which is not surprising as the data sample is not normally distributed --- that is, it has a large volume of higher glucose values, not unexpected, since most physicians and housestaff will request more frequent BG testing in diabetic patients or in those who are newly discovered to be hyperglycemic. This will effectively increase the mean. Here are the mean and median BGs for the 3 units of analysis. As you can see, there is little difference in the means. Also, the medians are smaller, which is not surprising as the data sample is not normally distributed --- that is, it has a large volume of higher glucose values, not unexpected, since most physicians and housestaff will request more frequent BG testing in diabetic patients or in those who are newly discovered to be hyperglycemic. This will effectively increase the mean.

30. The proportions of BG within the pre-specified range are also similar between the models.The proportions of BG within the pre-specified range are also similar between the models.

31. Here, in the % Hyperglycemic events, we see greater differences between the models, with the PATIENT showing the highest %, followed by PATIENT DAY, and then WARD. These differences aren’t surprising, since, for the PATEINT, a single hyperglycemic episode would characterize the whole patient, and similarly, an entire patient day would be characterized as “hyperglycemic” by a single episode in a day. While on the WARD, the hyperglycemic value would have the same weight as any other BG value - and they would be “diluted” out by the much larger number of BGs less than 300 mg/dl. Despite the simplicity of this notion, it has not yet been addressed in the emerging inpatient diabetes literature. Moreover, it is an important point. For instance a patient care unit with a hyperglycemia event rate of 12.8% sounds much better than one with an event rate of 39.0%.Here, in the % Hyperglycemic events, we see greater differences between the models, with the PATIENT showing the highest %, followed by PATIENT DAY, and then WARD. These differences aren’t surprising, since, for the PATEINT, a single hyperglycemic episode would characterize the whole patient, and similarly, an entire patient day would be characterized as “hyperglycemic” by a single episode in a day. While on the WARD, the hyperglycemic value would have the same weight as any other BG value - and they would be “diluted” out by the much larger number of BGs less than 300 mg/dl. Despite the simplicity of this notion, it has not yet been addressed in the emerging inpatient diabetes literature. Moreover, it is an important point. For instance a patient care unit with a hyperglycemia event rate of 12.8% sounds much better than one with an event rate of 39.0%.

32. The pattern of difference between the units of analysis is the same here with the hypoglycemic events showing a greater effect in the PATIENT and PATIENT DAY units of analysis than the WARD model. Again, a patient care unit with a hypoglycemia event rate of 1.7% sounds like relatively good “quality”, whereas one which is nearly 5-fold higher, at 7.6% suggests poor “quality.”The pattern of difference between the units of analysis is the same here with the hypoglycemic events showing a greater effect in the PATIENT and PATIENT DAY units of analysis than the WARD model. Again, a patient care unit with a hypoglycemia event rate of 1.7% sounds like relatively good “quality”, whereas one which is nearly 5-fold higher, at 7.6% suggests poor “quality.”

33. Metrics for mean BG, median BG & the % “in target range” are similar for all three analytical models. There were substantial differences between the models for % hyperglycemia and % hypoglycemia. ‘WARD’ model has lowest % ‘PATIENT STAY’ model has highest % ‘PATIENT DAY’ model is intermediate In summary, the global measures of Mean, Median and % in Range were similar for all 3 Units of Analysis. However, there were substantial differences between the models in % hyper- and hypoglycemia, which is not surprising as the PATIENT view highly weights a single hyper- or hypoglycemic event in a single patient, while the WARD view dilutes the effect of these events. The PATEINT DAY view is between these two.In summary, the global measures of Mean, Median and % in Range were similar for all 3 Units of Analysis. However, there were substantial differences between the models in % hyper- and hypoglycemia, which is not surprising as the PATIENT view highly weights a single hyper- or hypoglycemic event in a single patient, while the WARD view dilutes the effect of these events. The PATEINT DAY view is between these two.

34. Addition of venous plasma lab glucose measurements to fingerstick data Slight reduction in mean glucose values, but not clinically meaningful Deletion of 1st hospital day of blood glucose Slight reduction in mean glucose values, but not clinically meaningful Applying glucometrics to the ICU (‘gold standard’ with IV insulin infusion) - the realistic maximum % of patient days within target range is probably ~ 80% Obviously, while glucose control in hospitalized patients is important, especially in the ICU, measures directed solely at numerical performance may not provide an adequate assessment of true “quality.” Here are other features that might characterize the quality of a hospital’s inpatient glucose / diabetes management, which were not addressed by our investigation.Obviously, while glucose control in hospitalized patients is important, especially in the ICU, measures directed solely at numerical performance may not provide an adequate assessment of true “quality.” Here are other features that might characterize the quality of a hospital’s inpatient glucose / diabetes management, which were not addressed by our investigation.

35. Glucometrics are useful intermediate outcomes measures of inpatient hyperglycemia management. Perception of performance & quality may depend upon the unit of analysis All 3 Units of Analysis provide useful information ‘WARD’ model is the simplest; may be most useful in operational analyses. ‘PATIENT STAY’ model perhaps most useful to consumers (& risk management). ‘PATIENT DAY’ model may be the most actionable by providers. In conclusion, it’s clear that BG management is important to inpatient clinical outcomes. We believe our analysis demonstrates that BG metrics are useful hospital quality measures for BG management, as the glucose meter data are readily obtainable for analysis and provide measures of both Quality (BG control) and Safety (hypoglycemic events). To make use of these measures, health care institutions must establish rigorous methods for collection and analysis of BG data. It’s important to note that the perception of performance & quality for Hyper and Hypoglycemia depends largely upon the Unit of Analysis chosen for study. In conclusion, it’s clear that BG management is important to inpatient clinical outcomes. We believe our analysis demonstrates that BG metrics are useful hospital quality measures for BG management, as the glucose meter data are readily obtainable for analysis and provide measures of both Quality (BG control) and Safety (hypoglycemic events). To make use of these measures, health care institutions must establish rigorous methods for collection and analysis of BG data. It’s important to note that the perception of performance & quality for Hyper and Hypoglycemia depends largely upon the Unit of Analysis chosen for study.

36. Inpatient Diabetes Management Team: Impact on Glucometrics (Before vs. After) We then analyzed the BG for the IDMT vs. Non-IDMT groups, synchronizing for the hospital day on which the IDMT was consulted. The reduction in mean BG after consultation was -49.5 mg/dl in IDMT patients (which was a statistically significant change) vs. -16.4 mg/dl for the corresponding hospital days of matched non-IDMT patients (not significant). Significantly more IDMT patients moved into the target BG range than did non-IDMT patients, and there was a corresponding significant reduction Hyperglycemia among Team pts (but not among Non-Team pts) while there was no significant increase in Hypoglycemia.We then analyzed the BG for the IDMT vs. Non-IDMT groups, synchronizing for the hospital day on which the IDMT was consulted. The reduction in mean BG after consultation was -49.5 mg/dl in IDMT patients (which was a statistically significant change) vs. -16.4 mg/dl for the corresponding hospital days of matched non-IDMT patients (not significant). Significantly more IDMT patients moved into the target BG range than did non-IDMT patients, and there was a corresponding significant reduction Hyperglycemia among Team pts (but not among Non-Team pts) while there was no significant increase in Hypoglycemia.

37. These are the results comparing the IDMT cases’ Glucometrics Before vs. After Team consult. As you can see, the mean BG decreased by nearly 20% and the likelihood of having a Pt Day mean BG in the 70-149 target range more than doubled. The Pt Days with any marked hyperglycemia were reduced by more than half, without any substantial increase in Hypoglycemia. These are the results comparing the IDMT cases’ Glucometrics Before vs. After Team consult. As you can see, the mean BG decreased by nearly 20% and the likelihood of having a Pt Day mean BG in the 70-149 target range more than doubled. The Pt Days with any marked hyperglycemia were reduced by more than half, without any substantial increase in Hypoglycemia.

38. Other Proposed Metrics

39. “Time Average Glucose (TAG)”

40. Persistent Hyperglycemia: An Independent Predictor of AMI Outcomes

41. ‘Hyperglycemic Index’ As a Tool to Assess Glucose Control: A Retrospective Study 10 yr analysis in a 12-bed surgical ICU 1779 patients (LOS >4 days) 65,528 glucose values Obviously, while glucose control in hospitalized patients is important, especially in the ICU, measures directed solely at numerical performance may not provide an adequate assessment of true “quality.” Here are other features that might characterize the quality of a hospital’s inpatient glucose / diabetes management, which were not addressed by our investigation.Obviously, while glucose control in hospitalized patients is important, especially in the ICU, measures directed solely at numerical performance may not provide an adequate assessment of true “quality.” Here are other features that might characterize the quality of a hospital’s inpatient glucose / diabetes management, which were not addressed by our investigation.

42. The “Hyperglycemic Index” (HGI) Calculation of the hyperglycemic index (HGI). All measured glucose values (black dots) and their corresponding sampling times are taken into account. The average over time is calculated for the area (shaded) under the glucose curve for hyperglycemic values only. The normal glucose range is indicated by the hatched area, with 6.0 mmol/L (dotted line) as the cutoff. HGI is the shaded area divided by the total length of stay. In this case, HGI is 0.73 mmol/L, as indicated by the dashed line. Note that normal or hypoglycemic measurements do not affect HGI, and thus, they do not falsely lower this index.

43. Receiver Operator Characteristic (ROC) Curves for Different Glucose Measures

45. Glucometrics Institution: Second Hospital Ward: Medical ICU Type: Adult Internal Medicine ICU 1/3/2005 – 2/1/2005 A patient’s glycemic control can be analyzed at three time resolutions. Individual glucoses measure control at the shortest interval, the time between samples. Mean glucose measures control for the longest interval, the entire hospital stay. A day’s mean glucose measures control for an intermediate interval, one day. Only this method has a fixed interval which allows better comparison of one patient to another. The figures below show frequency distributions; the dark bar on the x-axis between 70 and 149 shows a target or goal glucose range. Percentiles of the data are shown by the lines and dot over the histogram: 5–––25 •50 75–––95

46. Patient–stays with only a single glucose measurement were excluded from the analysis: Patient–stays in the data file 73 Patient–stays with only one glucose 9 (12.3%) Patient–stays after excluding those with a single glucose 64 Glucose sampling characteristics (In the histograms below, data outliers beyond the 95th percentile are not shown.)

47. “Glucometrics”: Assessing Quality in Inpatient Glycemic Management SUMMARY “Quality” in inpatient glucose management needs to be better defined. Achieving it requires efforts in 4 spheres: prioritization; education; patient care; and metrics. Measures of inpatient glucose management are dependent on the analytical methods employed. It is important for the diabetes community, hospitals, clinical investigators & the QI experts work together to better define & validate standardized “glucometrics” which are meaningful, fair, and actionable.

  • Login