Implementing a Standardized Order Set for Community-Acquired Pneumonia: Impact on Mortality and Cost

Implementing a Standardized Order Set for Community-Acquired Pneumonia: Impact on Mortality and Cost

The Joint Commission Journal on Quality and Patient Safety Evidence-Based Medicine Implementing a Standardized Order Set for Community-Acquired Pneum...

367KB Sizes 0 Downloads 3 Views

The Joint Commission Journal on Quality and Patient Safety Evidence-Based Medicine

Implementing a Standardized Order Set for Community-Acquired Pneumonia: Impact on Mortality and Cost Neil S. Fleming, Ph.D.; Gerald Ogola, M.S., M.P.H.; David J. Ballard, M.D., M.S.P.H., Ph.D.

T

ools to standardize care such as order sets, templates, or protocols can improve compliance with recommended processes of care1–5 and patient outcomes.2,6–8 The impact of such tools on resource use appears more variable, depending in part on the clinical area or type of care targeted, with reductions in length of stay and overall charges being seen more consistently for inpatient surgical procedures9–16 than for conditions requiring inpatient medical management.14,16–18 Baylor Health Care System (BHCS) implemented a standardized order set for adult pneumonia in all its acute care hospitals in 2006. Preliminary evaluation based on the first 18 months of data showed results consistent with other studies examining the impact of similar tools on quality of care measures, patient outcomes, and financial indicators for conditions requiring in-hospital medical management: significant reduction in in-hospital mortality and significant increase in compliance with recommended processes of care but nonsignificant impact on length of stay and direct cost.19 To further elucidate the impact of the order set on patient outcomes, we now examine the 30-day mortality for adult pneumonia patients to determine whether the benefit seen in inpatient mortality was sustained following discharge; we also extend the cost-effectiveness analysis.

Methods STUDY SETTING BHCS is a not-for-profit, multihospital system in Dallas-Fort Worth, Texas, incorporating 20 owned, leased, affiliated, and short-stay hospitals with more than 103,000 admissions per year. Only the 8 acute care BHCS hospitals, where most of the patients with community-acquired pneumonia (CAP) are treated, were included in this study.

ADULT PNEUMONIA ORDER SET The adult pneumonia order set,19 shown in Appendix 1 (available in online article), was deployed systemwide via the physician portal in March 2006. At this time, “order set use” 414

August 2009

Article-at-a-Glance Background: Order sets have shown some success in

improving compliance with clinical guidelines, as well as patient and financial outcomes. Baylor Health Care System (BHCS) deployed a standardized adult pneumonia order set throughout its eight acute care hospitals in 2006. Methods: All non–comfort care adult patients admitted with community-acquired pneumonia who met The Joint Commission definition of pneumonia and were discharged from an acute care BHCS hospital for a 30-month period (March 1, 2006–August 31, 2008) were included. Mortality, core measures compliance, and direct cost were compared for patients who did and did not receive the order set. Results: Some 4,454 patients met study inclusion criteria. Significant variation in use between hospitals, however, persisted. Unadjusted analysis showed significant reductions in inhospital mortality, 30-day mortality, and direct cost and a significant increase in core measures compliance. Following risk adjustment, the difference in core measures compliance was retained (relative risk [95% confidence interval (C.I.)] 1.08 [1.03, 1.12]). Inhospital mortality and 30-day mortality reductions both approached significance (hazard ratios [95% C.I.] of 0.73 [0.51,1.02] and 0.79 [0.62, 1.00], respectively). Mean (standard error) benefits of order set use in in-hospital mortality and costs were estimated at 1.67 (0.62)% and $383 (207). The incremental cost-effectiveness ratio point estimate was –$22,882 per life saved, with an upper 95% confidence limit of $1,278 per life saved. Discussion: Widespread adoption of the order set was achieved, with use consistently at or above 75% across all BHCS acute care hospitals since February 2007. The reductions in mortality observed with order set use, in combination with the favorable estimate of cost-effectiveness, make standardized evidence-based order sets an attractive improvement methodology for improving quality of pneumonia care.

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

The Joint Commission Journal on Quality and Patient Safety was made a required field in MIDAS—the integrated outcomes, resource, and case management system used at all BHCS hospitals—for pneumonia patients, facilitating tracking of order set use. Strategies to increase awareness and encourage use of the order set included the following: ■ A high-profile awareness campaign, presented to the Best Care Committee and available to frontline care providers through the BHCS intranet ■ Just-in-time training provided to nursing units at some BHCS acute care hospitals ■ Incorporation in the order entry system in use at Baylor University Medical Center ■ Academic detailing by physician champions selected among physicians across BHCS hospitals, who were identified as highly regarded for their clinical acumen and interested in quality of care leadership roles and who had completed the 60hour BHCS rapid-cycle clinical process improvement curriculum known as Accelerating Best Care (ABC)–Baylor. The academic detailing tactics included reviewing the evidence basis for the pneumonia order set elements with individual physicians, providing individual physician-level feedback regarding use of the pneumonia order set, and individual discussions regarding specific rapid-cycle interventions to be tested to improve pneumonia order set use.

PATIENTS The study sample consisted of all adults (> 18 years of age) discharged from one of the eight BHCS acute care hospitals between March 1, 2006, and August 31, 2008, following treatment for community-acquired pneumonia and defined as eligible for one of the following Joint Commission pneumonia core measures20: ■ PN-5b Initial Antibiotic Received Within 4 Hours Of Hospital Arrival (or 6 Hours as of April 2007); ■ PN-6a Initial Antibiotic Selection For Community Acquired Pneumonia (CAP) In Immunocompetent–ICU Patient ■ PN-6b Initial Antibiotic Selection For Community Acquired Pneumonia (CAP) In Immunocompetent–non ICU patient Patients with physician orders for “comfort measures only” recorded in the admitting physician orders or note, consultation notes, emergency department record, history and physical, physician orders, or progress notes were excluded from the study.

OUTCOME MEASURES The primary outcome measures were differences in in-hos-

August 2009

Table 1. Joint Commission Core Measures for Pneumonia Included in the Pneumonia Composite Compliance Index PN-1 Oxygenation Assessment PN-2 Pneumococcal Vaccination PN-3b Blood Culture Before First Antibiotic PN-4 Adult Smoking Cessation Advice/Counseling PN-5b Initial Antibiotic Received Within 4 Hours Of Hospital Arrival* ■ PN-6a Initial Antibiotic Selection For Community Acquired Pneumonia (CAP) In Immunocompetent–ICU Patient ■ PN-6b Initial Antibiotic Selection For Community Acquired Pneumonia (CAP) In Immunocompetent–non ICU Patient ■ PN-7 Influenza Vaccination ■ ■ ■ ■ ■

* In April 2007, this measure was changed to “initial antibiotic received within 6 hours of hospital arrival.” Source: The Joint Commission: Pneumonia Core Measure Set, Last updated Apr. 2008. http://www.jointcommission.org/ PerformanceMeasurement/PerformanceMeasurement/Pneumonia+Core+ Measure+Set.htm (last accessed Jun. 23, 2009).

pital mortality; 30-day mortality; compliance with The Joint Commission core measures for pneumonia, as indicated by the pneumonia composite compliance index; and direct cost comparison between pneumonia patients treated with and without the BHCS standardized adult pneumonia order set. The core compliance index is based on eight of the pneumonia national quality measures, listed in Table 1 (above) and calculated as the proportion of pneumonia patients eligible for the above measures who receive all the measures for which they are eligible.21

DATA COLLECTION For each patient, age, sex, race/ethnicity, admitting BHCS hospital, compliance with Joint Commission core measures for pneumonia, and information on order set use (“BHCS order set” or “no BHCS order set”) were collected from MIDAS. Inpatient mortality and direct cost of care were determined from administrative data; 30-day mortality was determined from the Social Security Death Master File.22

STATISTICAL ANALYSIS The analysis considered only the first hospital admission during the study period to ensure that the statistical assumption of independence was met. Univariate analyses were conducted to examine the association or difference between order set use and patient characteristics or outcome measurements. Chisquare and Fisher exact tests were used for categorical characteristics and outcomes (sex, race, mortality, hospital facility, and

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

415

The Joint Commission Journal on Quality and Patient Safety core measures compliance). For ordinal variables (All Patient Refined Diagnosis Related Group [APR DRG] risk of mortality, severity of illness, Greenfield comorbidity,23 and month of discharge), mean score tests, which account for the category rankings, were used. Two-sample t-tests were used on continuous characteristics and outcomes that did not violate normality assumption. Cost data were highly skewed; therefore, robust tests were used. Regression Model. A multivariable Cox proportional hazard regression model was used to determine the adjusted effect of order set use on in-hospital mortality and 30-day postadmission mortality. Potential covariates to be adjusted for included age, sex, race, physician specialty (hospitalist versus nonhospitalist), payor type, discharged time (month, quarter, or year), APR DRG risk of mortality, Greenfield comorbidity, and admission source (emergency room versus other). A Cox model was fitted with all potential covariates, and then a manual backward elimination process was used to determine covariates whose effects were confounded with order set use. Covariates whose inclusion or exclusion in the multivariable model did not change the effect of order set on mortality were excluded from the model—admission source, primary diagnosis (pneumonia versus other), and physician specialty. The final model consisted of order set use, age, sex, race, discharge year, facility, APR DRG risk of mortality, and payor type. Hospital length of stay was considered as survival time in the in-hospital mortality model. Patients who survived beyond discharge were considered as censored observations in the analysis. Survival time for postadmission mortality was determined by the number of days from date of hospital admission to date of death on the Social Security Death Master File. Event occurrence was defined as death within 30-days from date of hospital admission. All patients surviving beyond 30 days were considered as censored observations. Core Measure Compliance. The adjusted risk ratio for core measures compliance with order set use was estimated via a multivariable log-binomial model.24 Logistic regression was not appropriate because core measures compliance was a frequently occurring outcome, and odds ratios overestimate risk in such circumstances. We used the same modeling approach as in the Cox model as described, resulting in the same covariates being considered for adjustment of the risk ratio. Cost-Effectiveness Analysis and Mortality. The cost-effectiveness analysis used hospital costs based on direct patient care for the inpatient episode, adjusted for inflation using the Medical Component of the Computer Price Index based on the 2008 base year,25 and in-hospital mortality for consistency in 416

August 2009

the time period and portion of care examined between mortality and cost data. The methodology employed was similar to that used by the National Emphysema Treatment Trial Research Group,26 which incorporates the cost of the intervention (in this case, deployment and use of the standardized order set) within the total medical treatment cost. Both costs and outcomes were modeled to compare patients treated with the BHCS order set with those treated without an order set, risk adjusted on the basis of APR DRG severity levels, comorbidities, and other impacting variables such as patient demographics (for example, age, sex). Covariates included in the models were age, sex, race, physician specialty (hospitalist versus other), primary diagnosis (pneumonia versus other), Greenfield’s comorbidity, severity of illness, payor type, admission source (emergency room versus other), and time (discharge quarter). To correctly model the nonnormal in-hospital mortality and cost data, the binary clinical outcome of mortality was analyzed using multiple logistic regression and the cost data with the log link function and multiple regression with a gamma distribution.27 Recycled predictions were used to account for and balance baseline differences between patients who did and did not receive the pneumonia order set.27 Specifically, mortality and costs were predicted from the modeled equations based on two scenarios: (1) every patient received the pneumonia order set and (2) every patient did not receive the pneumonia order set. The difference between these two predictions constitutes the predicted mean differences in in-hospital mortality and cost. We generated 1,000 bootstraps of this process to estimate (1) the mean differences in in-hospital mortality and costs and (2) the standard errors of these statistics. This approach permits estimation of incremental cost-effectiveness ratios (ICERs) and surrounding confidence intervals for the pneumonia order set. Specifically, Fieller’s Method can be applied to the bootstrapped results to generate the bounds of the confidence interval for the ICER, accounting for the correlation between clinical outcomes and costs.27 ICERs can be interpreted as the additional (marginal) cost of reducing incidence of a negative outcome (for example, the reduction of one death due to pneumonia). Potential Life Years. Potential life years saved per patient saved were calculated by estimating the number of life years lost for each person who died in the study. This was a two-part process: (1) Life expectancy was projected on the basis of the age and sex of the patient according to actuarial tables for patients (based on 2004 rates28), discounted based on the fiveyear survival reported for pneumonia patients,29 and (2) the difference was calculated between projected life expectancy and the age of the patient at time of death.

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

The Joint Commission Journal on Quality and Patient Safety Identification and Exclusion of Patients

Table 2. Use of the Baylor Health Care System (BHCS) Adult Pneumonia Standardized Order Set by Facility for Adult Pneumonia Patients Admitted to BHCS Acute Care Hospitals, March 1, 2006–August 31, 2008 Facility Hospital A Hospital B Hospital C Hospital D Hospital E Hospital F Hospital G Hospital H

N 533 200 1358 675 531 492 334 331

Order set BHCS n (%) None n (%) 370 (69) 163(31) 161 (80) 39 (20) 920 (68) 438 (32) 542 (80) 133 (20) 345 (65) 186 (35) 297 (60) 195 (40) 300 (90) 34 (10) 307 (93) 24 (7)

p Value* < .01

* Based on chi-square test.

(Table 2, above); the mean order set use was 76%. Patients in higher APR DRG risk-of-mortality and severity-of-illnesses classes were significantly less likely to receive the order set when variation in order set use by APR DRG classes was examined using mean score statistics that take into account the ordinal nature of these categorizations (p < .01 for both analyses).

MORTALITY Figure 1. This flowchart illustrates the identification and exclusion of patients included in the evaluation of the Baylor Health Care System (BCHS) Adult Pneumonia Order Set.

All analyses were conducted using SAS 9.2 (SAS Institute; Cary, North Carolina).

Results ORDER SET USE Between March 1, 2006, and August 31, 2008, 8,689 adult patients were admitted to BHCS hospitals with pneumonia. Of these, 4,454 met the study inclusion criteria (Figure 1, above). Order set use varied significantly by month (p < .001): “BHCS order set” use increased from 35% in March 2006 to 76% in August 2008. The average patient age was 66 ± 18 years. Approximately half (55%) of the included patients were women, and 76% were white. No significant variation in order set use by age, sex, or race was observed. Despite a very focused standardized approach to order set deployment across all hospitals, dramatic variation in use of the BHCS standardized order set was seen between hospitals (p < .01), with use ranging from 60% of pneumonia patients at one hospital to 93% at another

August 2009

A total of 168 patients (3.8%) died during their hospital stay. A significantly lower crude in-hospital death rate was observed among patients treated with the BHCS order set compared with those treated with no order set (3.0% versus 5.9%, p < .01; Table 3, page 418). An overall mortality of 7.6% (338 of 4,454) was observed at 30 days postadmission. A significantly lower crude 30-day postadmission mortality rate was observed among patients treated with order set compared with those treated with no order set (7.0% versus 9.1%, p = .02; Table 3). There were an additional 177 (4.1%) deaths among the 4,286 patients discharged alive that occurred within 30 days from admission date. Among the patients discharged alive, a nonsignificantly higher crude death rate was observed among patients treated with order set compared with those treated with none (4.2% [133 of 3,146] versus 3.9% [44 of 1,140], p = .59).

CORE MEASURES COMPLIANCE Table 3 also shows a statistically significant higher unadjusted rate of core measures compliance with order set use (79.8% versus 71.4%, p < .01), and a statistically significant lower unadjusted direct cost ($6,305 ± 8,069 versus $7,949 ± 12,196, p = .02).

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

417

The Joint Commission Journal on Quality and Patient Safety Table 3. Comparison of Quality of Care and Financial Indicators for Adult Pneumonia Patients Admitted to Baylor Health Care System (BHCS) Acute Care Hospitals, March 1, 2006-August 31, 2008 by Use of the BHCS Order Set (Unadjusted Results)* Order Set In-hospital mortality n (%) 30-day mortality n (%) Pneumonia core measures compliance n (%) Direct cost ($) mean (S.D.)

All 168 (3.8) 338 (7.6) 3,453 (77.5) 6,753 (9,401)

BHCS 96 (3.0) 228 (7.0) 2,588 (79.8) 6,305 (8,069)

None 72 (5.9) 110 (9.1) 865 (71.4) 7,949 (12,196)

P Value* < .01† < .02† < .01† .01‡

* S.D., standard deviation. † Based on chi-square test. ‡ Based on robust ANOVA test.

RISK-ADJUSTED EFFECTS OF ORDER SET USE ON MORTALITY, CORE MEASURE COMPLIANCE, AND COSTS Table 4 (page 419) shows the unadjusted and adjusted effect of BHCS order set use on mortality and core measures compliance. Adjusted analyses show a reduction in both in-hospital mortality and 30-day postadmission mortality. The adjusted decrease in hazard of either in-hospital or 30-day postadmission mortality bordered on statistical significance. The significant increase in core measures compliance observed in the unadjusted analysis was retained in the adjusted analysis. A total of 3,882 patients (2,861 in the BHCS order-set group and 1,021 in no-order-set group) survived beyond 90 days and were considered in the long-term survival analysis. Among the 2,861 patients treated with the order set, there were 396 (13.8%) deaths, whereas 210 (20.6%) deaths were observed for the no-order-set group of patients. Comparing mortality among patients of the same age group who survived beyond 90 days in the order-set and no-order-set group and that of the United States population from data derived from life tables30 showed statistically significant higher survival rate among patients treated with the order set (p < .01 based on log-rank test). Mean difference (standard error) in in-hospital mortality and costs were estimated at 1.67 (0.62)% and $383 (207) respectively, with both showing a benefit with order set use. The correlation between cost and in-hospital mortality was 0.138. Potential life years saved, based on adjusted lifeexpectancy calculations for all patients in the study population who died, was estimated at 12 years per patient. From the estimates of mortality and cost differences, the ICER point estimate was calculated as –$22,882 per additional life saved. Application of Fieller’s method to the estimates from the bootstrap replications gave an upper 95% confidence limit for the ICER of $1,278 per life. 418

August 2009

Discussion We investigated the impact of implementation of a standardized order set for pneumonia across a multihospital system on quality, cost, and outcomes of care for patients admitted with pneumonia during a 30-month period. During this time, widespread adoption of the order set was achieved, with use consistently at or above 75% across all BHCS acute care hospitals since February 2007. Significant variation in use between hospitals, however, persists. Variation was also observed with respect to some patient characteristics, with order set use being less likely for patients in higher APR DRG risk-of-mortality and severity-of-illnesses classes. Unadjusted results showed significant improvements with order set use in pneumonia core measures compliance, in-hospital mortality, 30-day mortality, and direct cost. Following adjustment for patient and provider characteristics, severity of illness, discharge month, and hospital, the only measure to retain statistical significance was core measures compliance, although the reductions in in-hospital and 30-day mortality bordered on significance. Mean (standard error) benefits of order set use in in-hospital mortality and costs were estimated at 1.67 (0.62)% and $383 (207) respectively. The ICER point estimate was calculated at –$22,882 per life saved, with an upper 95% confidence limit of $1,278 per life saved. The > 75% order set use rate observed is substantially higher than adoption rates previously reported for pneumonia order sets/clinical pathways, which have been below 30%.6,31,32 Our observation that use of the order set was associated with decreased mortality, albeit borderline statistically significant, was consistent with previous studies examining the impact of order set use on mortality, both for pneumonia8 and other conditions.7 Similarly, the finding that pneumonia core measures compliance improved was consistent with the results of previous studies that examined the impact of tools to increase the

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

The Joint Commission Journal on Quality and Patient Safety Table 4. The Unadjusted and Adjusted Effect of Baylor Health Care System (BHCS) Order Set Use for Quality Measures for Adult Pneumonia Patients Discharged from BHCS Acute Care Hospitals, March 1, 2006–August 31, 2008* Outcome In-hospital mortality, H.R. (95% C.I.) 30-day postadmission mortality, H.R. (95% C.I.) Core measures compliance, R.R. (95% C.I.)

Unadjusted 0.62 (0.46; 0.85) 0.76 (0.61; 0.96) 1.12 (1.08; 1.16)

Adjusted† 0.73 (0.51; 1.02) 0.79 (0.62; 1.00) 1.08 (1.03; 1.12)

* H.R., hazard ratio; R.R., relative risk; C.I., confidence interval. † Adjusted for age, sex, race, hospital facility, payor type, discharge year, and All Patient Refined Diagnosis Related Group (APR DRG) risk of mortality.

standardization of care—such as order sets and protocols—on process of care quality indicators.1,3,4,7,33,34 The mean differences in cost and in-hospital mortality estimated to calculate the ICER indicated reductions in both these measures with order set use. The resulting ICER point estimate was thus negative, and so careful attention must be paid to the interpretation of the magnitude.27 Looking simply at the ICER point estimate, we might be tempted to conclude that to maximize the mortality reduction—given the equivalence of cost or even possible reduction in cost with order set use—the order set should be uniformly adopted. This, however, does not consider the sampling distribution of the ICER; implementation of the order set could lead to different results. It is important to look at the 95% confidence interval surrounding the point estimate. The upper limit of the confidence interval defines the lowest maximum acceptable cost-effectiveness ratio for the mortality reduction. The upper 95% confidence limit here of $1,278 per life saved observed here falls well under the conventional limit of $50,000 per life year used in identifying an intervention as “cost-effective” or not. Specifically, the average years saved per life with order set use was estimated at 12 years, which represents approximately 15% of the average adjusted life expectancy of 83 years for our study population, as calculated based on United States actuarial tables.35

LIMITATIONS Because this was an observational study as opposed to a randomized trial, it is possible that order set use was influenced by a patient’s characteristics, potentially masking or exaggerating the impact of order set use. To account for the differences between the patients that did and did not receive the order set, we conducted analyses adjusted using a covariate approach, which also accounted for the variation in order set use by facility and time observed in this study. However, covariate risk adjustment is limited in that it does not ensure a balanced distribution of covariates among the study subjects.36 An additional aspect of this study design that cannot be dis-

August 2009

counted in interpreting results is the potential for contamination between study groups. Physicians and other clinicians may have been influenced by exposure to the order set in their care decisions even for patients to whom the order set was not applied. Such contamination would attenuate differences between the study groups, underestimating the impact of the order set.

INFRASTRUCTURE FOR ORDER SETS An important part of any quality improvement project is ensuring that the relevant infrastructure is in place to support the maintenance of gains achieved. At BHCS, this was incorporated within the overall system implemented for the development and deployment of systemwide order sets, which has been previously described.19 Leadership for each order set developed and deployed is provided by the physician champion (clinicians with dedicated BHCS–funded time for promoting quality improvement initiatives within BHCS) within whose clinical area the targeted condition or procedure falls. For example, the physician champion for hospitalist medicine oversees the pneumonia order set initiative. In addition, each order set is reviewed and updated annually by subspecialty teams, physician champions, and other physician leaders from the facilities and the physician design team (group of system leaders, nursing staff, and a pharmacist) to ensure that it aligns with the current evidence for best care. Any changes are reviewed by the BHCS pharmacy and therapeutics (P&T) and patient safety committees. Order set use and related outcomes are monitored within the BHCS Center for Health Care Improvement and are reported to the relevant physician champions quarterly. Although standardized order set use is currently voluntary within BHCS hospitals, deployment is a preliminary step in the long-term goal of introducing computerized physician order entry. Use is strongly encouraged through town hall meetings, through one-on-one academic detailing by physician champions, at medical department and section meetings, and through other staff education. The adult pneumonia order set was the

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

419

The Joint Commission Journal on Quality and Patient Safety first of many standardized order sets BHCS is introducing, and the process for developing, implementing, and monitoring order set was developed and refined in tandem with its introduction. To date, approximately 50 order sets have been deployed systemwide for conditions and procedures within internal medicine, cardiology, obstetrics, neurology, gastroenterology, cardiothoracic surgery, transplant surgery, general surgery, and orthopedic surgery, as well as for more generally applicable targets such as venous thromboembolism and glucose control. An order set for adult admission to the ICU for sepsis is currently being piloted at two hospitals.

Summary For the pneumonia order set, the results show meaningful improvements in patient outcomes associated with the implementation of a standardized order set throughout a health care system. The reductions in in-hospital and 30-day mortality with order set use bordered on statistical significance. The substantial but steady increase in use of the order set during the 30month study period suggests that health care organizations considering similar initiatives to improve quality of care should anticipate a period of several months before use of the order sets is sufficiently integrated into clinical practice to achieve detectable changes. Order set use was also associated with lower costs, and evaluation of the cost-effectiveness of the order set was favorable with the upper level of the ICER 95% confidence interval, indicating the mortality reduction can be achieved at a cost of only $1,278 per life saved. Overall, the BHCS experience suggests that implementation of such an order set is an attractive improvement methodology for hospitals seeking to raise the quality of their pneumonia care. J The authors thank Donald Kennerly, M.D., Ph.D., for advice regarding clinical severity/mortality modeling; Roger Khetan, M.D., and Jeff Kerr, M.D., for their leadership of the pneumonia order set development and implementation; Andrew Masica, M.D., M.S.C.I., for advice related to the cost-effectiveness analysis; Henry Glick, Ph.D., the University of Pennsylvania School of Medicine, for his help in interpreting the results regarding the sampling uncertainty of the incremental costeffectiveness ratio; Raaj Mehta, for work related to this project during his summer internship; and Briget da Graca, M.S., E.L.S., for background research, writing, and editorial assistance.

Neil S. Fleming, Ph.D., is Vice President for Health Care Research, Institute for Health Care Research and Improvement, Baylor Health Care System, Dallas; Gerald Ogola, M.S., M.P.H., is Biostatistician; and David J. Ballard, M.D., M.S.P.H., Ph.D., is Senior Vice President and Chief Quality Officer. Please address correspondence to David J. Ballard, [email protected].

420

August 2009

Online-Only Content

8

See the online version of this article for

Appendix 1. Baylor Health Care System Standardized Adult Pneumonia Order Set

References 1. Biviano A.B., et al.: Usefulness of an acute coronary syndrome pathway to improve adherence to secondary prevention guidelines. Am J Cardiol 91:1248–1250, May 15, 2003. 2. Fonarow G.C., et al.: Improved treatment of coronary heart disease by implementation of a Cardiac Hospitalization Atherosclerosis Management Program (CHAMP). Am J Cardiol 87:819–822, Apr. 1, 2001. 3. Mehta R.H., et al.: Improving quality of care for acute myocardial infarction: The Guidelines Applied in Practice (GAP) initiative. JAMA 287:1269–1276, Mar. 13, 2002. 4. Ozdas A., et al.: Integrating “best of care” protocols into clinicians’ workflow via care provider order entry: Impact on quality-of-care indicators for acute myocardial infarction. J Am Med Inform Assoc 13:188–196, Mar.–Apr. 2006. 5. California Acute Stroke Pilot Registry Investigators: The impact of standardized stroke orders on adherence to best practices. Neurology 65:360–365, Aug. 9, 2005. 6. Hauck L.D., et al.: Clinical pathway care improves outcomes among patients hospitalized for community-acquired pneumonia. Ann Epidemiol 14:669–675, Oct. 2004. 7. Micek S.T., et al.: Before-after study of a standardized hospital order set for the management of septic shock. Crit Care Med 34:2707–2713, Nov. 2006. 8. Dean N.C., et al.: Decreased mortality after implementation of a treatment guideline for community-acquired pneumonia. Am J Med 110:451–457, Apr. 15, 2001. 9. Bradshaw B.G., et al.: Standardized perioperative care protocols and reduced length of stay after colon surgery. J Am Coll Surg 186:501–506, May 1998. 10. Greenwald J.A., et al.: Standardization of surgeon-controlled variables: Impact on outcome in patients with acute cholecystitis. Ann Surg 231:339–344, Mar. 2000. 11. Hanna E., et al.: Development and implementation of a clinical pathway for patients undergoing total laryngectomy: Impact on cost and quality of care. Arch Otolaryngol Head Neck Surg 125:1247–1251, Nov. 1999. 12. Jordan P., et al.: Decreasing process variation in the care of carotid endarterectomy patients. Top Health Inf Manage 22:24–34, Nov. 2001. 13. Kiyama T., et al.: (Clinical significance of a standardized clinical pathway in gastrectomy patients.) J Nippon Med Sch 70:263–269, Jun. 2003. 14. Panella M., et al.: Reducing clinical variations with clinical pathways: Do pathways work? Int J Qual Health Care 15:509–521, Dec. 2003. 15. Scranton P.E. Jr.: The cost effectiveness of streamlined care pathways and product standardization in total knee arthroplasty. J Arthroplasty 14:182–186, Feb. 1999. 16. Takegami K., et al.: Impact of a clinical pathway and standardization of treatment for acute appendicitis. Surg Today 33:336–341, 2003. 17. Halm E.A., et al.: Limited impact of a multicenter intervention to improve the quality and efficiency of pneumonia care. Chest 126:100–107, Jul. 2004. 18. Marrie T.J., et al.: A controlled trial of a critical pathway for treatment of community-acquired pneumonia. CAPITAL Study Investigators. Community-Acquired Pneumonia Intervention Trial Assessing Levofloxacin. JAMA 283:749–755, Feb. 9, 2000. 19. Ballard D.J., et al.: The impact of standardized order sets on quality and financial outcomes. Advances in Patient Safety: New Directions and Alternative Approaches, vol. 2. Rockville, MD: Agency for Healthcare Research and

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

The Joint Commission Journal on Quality and Patient Safety Quality. 2008. http://www.ahrq.gov/downloads/pub/advances2/vol2/ Advances-Ballard_12.pdf (last accessed Jun. 23, 2009). 20. The Joint Commission: Pneumonia Core Measure Set. Last updated Apr. 2008. http://www.jointcommission.org/PerformanceMeasurement/ PerformanceMeasurement/Pneumonia+Core+Measure+Set.htm (last accessed Jun. 23, 2009). 21. The Joint Commission: Current Specifications Manual for National Hospital Quality Measures. http://www.jointcommission.org/PerformanceMeasurement/ PerformanceMeasurement/Current+NHQM+Manual.htm (last accessed Jun. 23, 2009). 22. U.S. National Technical Reports Library: Social Security Administration: Death Master File. http://www.ntis.gov/products/ssa-dmf.aspx (last accessed Jun. 23, 2009). 23. Greenfield S., et al.: The importance of co-existent disease in the occurrence of postoperative complications and one-year recovery in patients undergoing total hip replacement. Comorbidity and outcomes after hip replacement. Med Care 31:141–154, Feb. 1993. 24. Spiegelman D., Hertzmark E.: Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol 162:199–200, Aug. 1, 2005. 25. Gold M.R., et al. (eds.): Cost-Effectiveness in Health and Medicine. New York City: Oxford University Press, 1996. 26. Ramsey S.D., et al.: Cost effectiveness of lung-volume-reduction surgery for patients with severe emphysema. N Engl J Med 348:2092–2102, May 22, 2003. 27. Glick H.A., et al.: Economic Evaluation in Clinical Trials. Oxford: Oxford University Press, 2007. 28. Social Security Administration: Period Life Table. 2004.

August 2009

http://www.ssa.gov/OACT/STATS/table4c6.html (last accessed Aug. 25, 2008). 29. Mortensen E.M., et al.: Assessment of mortality after long-term follow-up of patients with community-acquired pneumonia. Clin Infect Dis 37:1617–1624, Dec. 15, 2003. 30. University of California Berkeley (USA); Max Planck Institute for Demographic Research (Germany): The Human Mortality Database. http://www.mortality.org or http://www.humanmortality.de (last accessed Jun. 23, 2009). 31. McAlearney A.S., et al.: Utilization of evidence-based computerized order sets in pediatrics. Int J Med Inform, Sep. 15, 2005. 32. Estrada C.A., et al.: Judging the effectiveness of clinical pathways for pneumonia: The role of risk adjustment. Eff Clin Pract 3:221–228, Sep.–Oct. 2000. 33. Mehta R.H., et al.: Quality improvement initiative and its impact on the management of patients with acute myocardial infarction. Arch Intern Med 160:3057–3062, Nov. 13, 2000. 34. Mehta R.H., et al.: Enhancing quality of care for acute myocardial infarction: Shifting the focus of improvement from key indicators to process of care and tool use: The American College of Cardiology Acute Myocardial Infarction Guidelines Applied in Practice Project in Michigan: Flint and Saginaw Expansion. J Am Coll Cardiol 43:2166–2173, Jun. 16, 2004. 35. National Center for Health Statistics: Life Expectancy. http://www.cdc.gov/nchs/fastats/lifexpec.htm (last accessed Jun. 23, 2009). 36. Dehejia R.H., Wahba S.: Causal effects in nonexperimental studies: Reevaluating the evaluation of training programs. Journal of the American Statistical Association 94:1053–1062, 1999.

Volume 35 Number 8

Copyright 2009 Joint Commission on Accreditation of Healthcare Organizations

421

5-Hole 1/4 1 3/8 c-to-c

BAYLOR HEALTH CARE SYSTEM PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET

CLINICAL

transformation

2-Hole 1/4 2 3/4 - 3-Hole 1/4 4 1/4

Admit Admit Status  Inpatient  Floor  Telemetry  ICU  Other  Observation Diagnosis 1. 2. 3. Allergies

Date:______________________________________________________________________________________________

  X To order/select an item by placing a mark in corresponding box  Attending Primary Care Physician Referring Physician

4. 5. 6. Reaction

Code Status  Full Code  Do Not Resuscitate/Do Not Intubate (allow natural death) Other ____________________________________________________________ Vital Signs  Vital Signs every _____  Neurological checks with vitals every _____ Deep Venous Thrombosis (DVT) Prophylaxis (*Cross out printed dose to write in another dosage, For example: renal dosing)  Enoxaparin (Lovenox) 40 mg (*____________) subcutaneous daily  Heparin 5,000 units subcutaneous every 8 hours  Sequential compression device AND Compression Hose Activity  Up ad lib  Bed Rest  Bed rest with bedside commode  Bed rest with bathroom privileges Nursing Orders Description   Weigh patient  

Pulse Oximetry (JCAHO Core Measure)   Initiate Oxygen protocol if oxygen saturation less than 92%  Strict Intake and Output  Insert Foley catheter  Discontinue Foley catheter after 48 hours   Notify Physician if: Temperature greater than 101.5 Fahrenheit Heart rate greater than 115 or less than 50 Other:

   

Out of bed to chair __________ (frequency) Ambulate with assistance ________ (frequency) Activity per Physical Therapy/Occupational Therapy Other Frequency  Admission   Admission   Continuous

Instructions  Daily  Intermittent

Every 8 hours

Respiratory Rate greater than 24 or less than 8 Blood pressure systolic greater than 180 or less than 90

Legend: BID=twice a day, BP=blood pressure, BPM=beats per minute, cm=centimeter, gm=gram, HR=heart rate, HS=hour of sleep, IM=intramuscular, IV=intravenous, IVPB=intravenous piggyback, JCAHO=Joint Commission on Accreditation of Healthcare Organizations, kg=kilogram, lbs=pounds, mcg=microgram, mg=milligram, ml=milliliter, NPO=nothing by mouth, OD=right eye, OS=left eye, oz=ounce, PO=by mouth, PR=per rectum, PRN=as needed, q=every, QID=four times a day, subQ=subcutaneous, TID=three times a day, UOM=units of measure

BAYLOR HEALTH CARE SYSTEM BHCS-49004 (Rev. 06/06)

PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET Page 1 of 5

5-Hole 1/4 1 3/8 c-to-c

CLINICAL

transformation

2-Hole 1/4 2 3/4 - 3-Hole 1/4 4 1/4

Diet  Regular  NPO Isolation Isolation Order:

BAYLOR HEALTH CARE SYSTEM PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET  NPO except Meds  Cardiac Diet

 Diabetic Diet

Laboratory Test Instructions  Stat  AM Draw Complete blood count Basic Metabolic Panel  Stat  AM Draw Complete Metabolic Panel  Stat  AM Draw Stat, prior to antibiotic administration.  Do not order test if Blood Cultures times 2  (JCAHO core measure) already done in Emergency Department.  Arterial Blood Gas  (JCAHO core measure) Stat, if Emergency Department or admission pulse oximetry less than 92%  Phosphorus  Stat  AM Draw  Magnesium  Stat  AM Draw  Urinalysis Nurse collect  Urine Culture and Sensitivity Nurse collect  Urine pregnancy test Nurse collect     

Diagnostic Testing Test  Chest x-ray, Portable  Chest x-ray, posterior/anterior and lateral IV Solutions Order  Saline Lock (flush per protocol)  0.9% Sodium Chloride  0.45% Sodium Chloride  Dextrose 5%  Dextrose 5% in 0.45% Sodium Chloride

Frequency  Now  Now

Reason for exam

Additive

Rate @ @ @ @

Medications PRN (*To prescribe a dose other than the pre-printed dose: cross out pre-printed dose and write alternate dose, For example: for renal dose considerations) Order Dose UOM Route Frequency Instructions  Magnesium Hydroxide Suspension 30 (*________) ml PO Once Daily PRN Constipation (Milk of Magnesia)  Antacid per Pharmacy 30 (*________) ml PO 3 times a day PRN Indigestion/ Heart Burn  Ondansetron (Zofran) 4 (*________) mg IV Every 6 hours PRN Nausea  Bisacodyl Suppository (Dulcolax) 10 (*________) mg PR Once Daily PRN Constipation  Docusate Sodium (Colace) 100 (*________) mg PO Every 12 hours PRN Constipation  Zolpidem (Ambien) 5 (*________) mg PO At bedtime PRN Insomnia  Acetaminophen (Tylenol) 650 (*________) mg PO Every 4 hours PRN Pain 1-2/10 or fever or headache  Hydrocodone/acetaminophen (Norco) 5/325 (*________) 1 tablet PO Every 4 hours PRN Pain 3-5/10  Acetaminophen/propoxyphene (Darvocet N) 650/100 (*_______) 1 tablet PO Every 4 hours PRN Pain 3-5/10 *Avoid use in patients greater than 65 years of age  Other: Legend: BID=twice a day, BP=blood pressure, BPM=beats per minute, cm=centimeter, gm=gram, HR=heart rate, HS=hour of sleep, IM=intramuscular, IV=intravenous, IVPB=intravenous piggyback, JCAHO=Joint Commission on Accreditation of Healthcare Organizations, kg=kilogram, lbs=pounds, mcg=microgram, mg=milligram, ml=milliliter, NPO=nothing by mouth, OD=right eye, OS=left eye, oz=ounce, PO=by mouth, PR=per rectum, PRN=as needed, q=every, QID=four times a day, subQ=subcutaneous, TID=three times a day, UOM=units of measure

BAYLOR HEALTH CARE SYSTEM BHCS-49004 (Rev. 06/06)

PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET Page 2 of 5

5-Hole 1/4 1 3/8 c-to-c

CLINICAL

transformation

BAYLOR HEALTH CARE SYSTEM PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET

2-Hole 1/4 2 3/4 - 3-Hole 1/4 4 1/4

Antibiotic Medications (* To prescribe a dose other than the pre-printed dose: cross out pre-printed dose and write alternate dose, For example: for renal dose considerations)  Pharmacist consult for drug dosing 

Non-Health Care Acquired Pneumonia (CAP), Non ICU Order  Azithromycin (Zithromax) AND  Ceftriaxone (Rocephin) OR  Levofloxacin (Levaquin) ONLY (consider renal dosing)

Dose UOM 500 (*________) mg 1 (*________) gm

Route Frequency IVPB Daily IVPB Daily

Instructions First dose Stat First dose Stat

500 (*________) mg

IVPB

First dose Stat

Daily

Severe Pneumonia, ICU (Defined as: Respiratory rate greater than 30 per minute; PaO2/FIO2 less than 250; Patient needing mechanical ventilation; Bilateral/multilobar infiltrates on chest x-ray; Increase in size of infiltrate on chest x-ray of 50% in first 48 hours; Systolic blood pressure less than 90; Diastolic blood pressure less than 60; Need for vasopressors for greater than 4 hours; Urine output less than 20ml per hour for greater than 4 hours; Acute renal insufficiency) Reminder: Quinolone cannot be given as monotherapy in severe, ICU pneumonia. Order Dose UOM Route Frequency  Ceftriaxone (Rocephin) 1 (*________) gm IVPB Daily AND Choose one of the following  Levofloxacin (Levaquin) 750 (*________) mg IVPB Daily (consider renal dosing)  Azithromycin (Zithromax) 500 (*________) mg IVPB Daily

Instructions First dose Stat

First dose Stat First dose Stat

Pneumonia with Pseudomonas Risk (Defined as: Recent hospitalization, especially ICU stay; Recent antibiotic exposure; Severe structural lung disease (bronchiectasis; cystic fibrosis); Gram negative rods in the sputum gram stain.) Order Choose one of the following

Dose

UOM

Route

Frequency

Instructions



Piperacillin sodium/tazobactam sodium (Zosyn) (consider renal dosing)

4.5

(*________) gm

IVPB

Every 6 hours

First dose Stat



Cefepime (Maxipime)

2

(*________) gm

IVPB

Every 12 hours

First dose Stat



Imipenem/cilastatin (Primaxin) (consider renal dosing)

500 (*________) mg

IVPB

Every 6 hours

First dose Stat

IVPB

Daily

First dose Stat

AND Choose One of the following 

Tobramycin (consider renal dosing)

5

(*________) mg/kg



Levofloxacin (Levaquin) (consider renal dosing)

750 (*________) mg

IVPB

Daily

First dose Stat



Gentamicin (consider renal dosing)

5

IVPB

Daily

First dose Stat

(*________) mg/kg

Legend: BID=twice a day, BP=blood pressure, BPM=beats per minute, cm=centimeter, gm=gram, HR=heart rate, HS=hour of sleep, IM=intramuscular, IV=intravenous, IVPB=intravenous piggyback, JCAHO=Joint Commission on Accreditation of Healthcare Organizations, kg=kilogram, lbs=pounds, mcg=microgram, mg=milligram, ml=milliliter, NPO=nothing by mouth, OD=right eye, OS=left eye, oz=ounce, PO=by mouth, PR=per rectum, PRN=as needed, q=every, QID=four times a day, subQ=subcutaneous, TID=three times a day, UOM=units of measure

BAYLOR HEALTH CARE SYSTEM BHCS-49004 (Rev. 06/06)

PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET Page 3 of 5

5-Hole 1/4 1 3/8 c-to-c

CLINICAL

transformation

BAYLOR HEALTH CARE SYSTEM PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET

2-Hole 1/4 2 3/4 - 3-Hole 1/4 4 1/4

Healthcare or Nursing Home Acquired Pneumonia ( Defined as: Antimicrobial therapy in preceding 90 days; Current hospitalization for 5 or more days; High frequency of antibiotic resistance in the community or in a specific hospital unit; Hospitalization for 2 days or more in the preceding 90 days; Home infusion therapy; Chronic dialysis within 30 days; Home wound care; Family member with multi-drug resistant pathogen; Immunosuppressive disease and/or therapy.) Order Dose UOM Route Frequency Instructions Choose one of the following 

Piperacillin sodium/tazobactam sodium (Zosyn) (consider renal dosing)

4.5

(*________) gm

IVPB

Every 6 hours

First dose Stat



Cefepime (Maxipime)

2

(*________) gm

IVPB

Every 12 hours

First dose Stat



Imipenem/cilastatin (Primaxin) (consider renal dosing)

500 (*________) mg

IVPB

Every 6 hours

First dose Stat

IVPB

Daily

First dose Stat

AND Choose One of the following 

Tobramycin (consider renal dosing)

5

(*________) mg/kg



Levofloxacin (Levaquin) (consider renal dosing)

750 (*________) mg

IVPB

Daily

First dose Stat



Gentamicin (consider renal dosing)

5

IVPB

Daily

First dose Stat

900 (*________) mg

IVPB

Every 8 hours

First dose Stat

Vancomycin (consider renal dosing)

1

(*________) gm

IVPB

Every 12 hours

First dose Stat

Linezolid (Zyvox)

600

mg

IVPB

Every 12 hours

First dose Stat

(*________) mg/kg

Additional Considerations IF ASPIRATION is suspected ADD Clindamycin (Cleocin)  only if not currently on Zosyn or Primaxin IF methicillin resistant staphylococcus aureus (MRSA) is suspected ADD 



OR 

Respiratory Care 

Oxygen Protocol



Other:

Education 

Conduct Smoking Cessation Counseling



Conduct Diabetes Education



Conduct Congestive Heart Failure Education

Legend: BID=twice a day, BP=blood pressure, BPM=beats per minute, cm=centimeter, gm=gram, HR=heart rate, HS=hour of sleep, IM=intramuscular, IV=intravenous, IVPB=intravenous piggyback, JCAHO=Joint Commission on Accreditation of Healthcare Organizations, kg=kilogram, lbs=pounds, mcg=microgram, mg=milligram, ml=milliliter, NPO=nothing by mouth, OD=right eye, OS=left eye, oz=ounce, PO=by mouth, PR=per rectum, PRN=as needed, q=every, QID=four times a day, subQ=subcutaneous, TID=three times a day, UOM=units of measure

BAYLOR HEALTH CARE SYSTEM BHCS-49004 (Rev. 06/06)

PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET Page 4 of 5

5-Hole 1/4 1 3/8 c-to-c

CLINICAL

transformation

BAYLOR HEALTH CARE SYSTEM PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET

2-Hole 1/4 2 3/4 - 3-Hole 1/4 4 1/4

Immunizations    

Pneumovax per protocol  (JCAHO core measure) Influenza Vaccine per protocol  (JCAHO core measure)

Departmental Referrals 

Physical Therapy Evaluation and Treatment

 Occupational Therapy

to evaluate activities of daily living

 Care 

Coordination

 Social

Work

 Speech Therapy

 Behavioral

 Nutritional

 Pastoral

Services

 Enterostomal Therapy

Health

Care

 Recreational Therapy

Patient Safety  Institute 

Fall Precautions

 Institute

Seizure Precautions

 Institute Aspiration

Precautions

Additional Orders

Physician Signature: Nurse Signature:

Provider #:

Date/Time: Date/Time:

am / pm am / pm

Legend: BID=twice a day, BP=blood pressure, BPM=beats per minute, cm=centimeter, gm=gram, HR=heart rate, HS=hour of sleep, IM=intramuscular, IV=intravenous, IVPB=intravenous piggyback, JCAHO=Joint Commission on Accreditation of Healthcare Organizations, kg=kilogram, lbs=pounds, mcg=microgram, mg=milligram, ml=milliliter, NPO=nothing by mouth, OD=right eye, OS=left eye, oz=ounce, PO=by mouth, PR=per rectum, PRN=as needed, q=every, QID=four times a day, subQ=subcutaneous, TID=three times a day, UOM=units of measure

BAYLOR HEALTH CARE SYSTEM BHCS-49004 (Rev. 06/06)

PHYSICIAN ORDERS ADULT PNEUMONIA ORDER SET Page 5 of 5