Using Information Technology to Improve Adult Immunization Delivery in an Integrated Urban Health System

The Joint Commission Journal on Quality and Patient Safety Information Technology

Using Information Technology to Improve Adult Immunization Delivery in an Integrated Urban Health System Carolyn J. Swenson, MSPH, MSN; Alicia Appel, MD; Moira Sheehan, MAS; Anne Hammer, BA, BSN, RN; Zita Fenner, BSN, RN, CNE; Stephanie Phibbs, MPH; Marjie Harbrecht, MD; Deborah S. Main, PhD

H

ealthy People 2010 set a goal of 90% influenza and pneumococcal immunization coverage among noninstitutionalized adults aged 65 years and older, and 60% coverage among eligible adults aged 18-64 years.1 Yet despite clear goals and evidence of the effectiveness of adult immunization,2,3 rates remain below expectations.4 Health care settings might often miss opportunities to identify patients who need immunizations.5-7 Standing order policies help improve adult immunization rates, but there is a need for additional strategies to ensure their full implementation.8–10 Computerized clinical decision support systems (CDSSs) that identify eligible patients and guide immunization delivery have the potential to improve immunization rates if they are built into a health care delivery system’s infrastructure.11,12 A CDSS might also decrease immunization costs by preventing duplicate services, provided that immunization documentation is accurate and consistently completed.13 Even with a CDSS in place, it is critical that it be available at the appropriate point in the patient care process, used by the appropriate members of the health care team, and integrated into the work flow so that it cannot be overlooked or ignored.14,15 Building an effective infrastructure is more likely to occur in health care organizations that demonstrate leadership support for innovation, encourage a high level of engagement of providers and other staff, value data and feedback to assess improvement efforts, and are able to spread and sustain change throughout a system.16–19 We have previously described a quality improvement (QI) initiative to improve adult pneumococcal, influenza, and tetanus immunization rates in a large, integrated, urban health care system, with development of a CDSS as a core component of the project.20,21 In this article, we summarize the project, describe subsequent modifications made to the CDSS (including expanding it to encompass additional immunizations), describe challenges and strategies used to address them, and note future systemwide changes that are expected to enhance the improvement and sustainability of adult immunization rates.

Article-at-a-Glance Background: Adult immunizations prevent morbidity and mortality yet coverage remains suboptimal, in part due to missed opportunities. Clinical decision support systems (CDSSs) can improve immunization rates when integrated into routine work flow, implemented wherever care is delivered, and used by staff who can act on the recommendation. Methods: An adult immunization improvement project was undertaken in a large integrated, safety-net health care system. A CDSS was developed to query patient records and identify patients eligible for pneumococcal, influenza, or tetanus immunization and then generate a statement that recommends immunization or indicates a previous refusal. A new agency policy authorized medical assistants and nurses in clinics, and nurses in the hospital, to use the CDSS as a standing order. Immunization delivery work flow was standardized, and staff received feedback on immunization rates. Results: The CDSS identified more patients than a typical paper standing order and can be easily modified to incorporate changes in vaccine indications. The intervention led to a 10% improvement in immunization rates in adults 65 years of age or older and in younger adults with diabetes or chronic obstructive pulmonary disease. Overall, the improvements were sustained beyond the project period. The CDSS was expanded to encompass additional vaccines. Conclusions: Interdepartmental collaboration was critical to identify needs, challenges, and solutions. Implementing the standing order policy in clinics and the hospital usually allowed immunizations to be taken out of the hands of clinicians. As an on-demand tool, CDSS must be used at each patient encounter to avoid missed opportunities. Staff retraining accompanied by ongoing assessment of immunization rates, work flow, and missed opportunities to immunize patients are critical to sustain and enhance improvements.

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

15

The Joint Commission Journal on Quality and Patient Safety Methods SETTING Denver Health and Hospital Authority (Denver Health) serves approximately 25% of all residents of the city and county of Denver, representing an estimated total population of 150,000, and is the state’s primary safety-net institution. Service delivery points include a 500-bed inpatient hospital designated as a Level 1 Trauma Center, an eight-clinic network of community health centers throughout the city, associated specialty clinics, the county public health department, a clinical detoxification unit, correctional care, a center for occupational health and safety, and a managed care clinic for employees of the city and county of Denver. Denver Health has a 150-year history of serving patients in the city and county of Denver and established one of the first community health centers in the United States in 1966. Adult immunizations are delivered in all its settings.

A QUALITY IMPROVEMENT PROJECT TO IMPROVE IMMUNIZATION RATES In 1996 Denver Health initiated a successful QI process to improve pediatric immunization.22 Central to that improvement process was the use of a systemwide immunization registry for accurate, complete documentation of pediatric immunizations and a “recommend function” to guide immunization delivery as part of well child care. Denver Health sought to apply lessons and methods from the pediatric immunization improvement experience to an adult improvement project, which included the use of technology, standing orders, and standardization of the immunization process. From 2005 through 2008 Denver Health partnered with HealthTeamWorks (formerly Colorado Clinical Guidelines Collaborative) and the University of Colorado Denver (UCD) to develop, implement, and evaluate an adult immunization QI project focused on pneumococcal and influenza immunizations. HealthTeamWorks (http://www.healthteamworks.org/) is a private, nonprofit coalition of multiple stakeholders working together to improve health care quality and efficiency throughout Colorado and throughout the United States, with a focus on developing and implementing evidence-based clinical guidelines through redesign at the clinical practice, community, and larger systems levels. UCD coordinated an evaluation of project processes and outcomes. An important initial step was to assemble an oversight team with representatives from management, information systems (IS), the medical staff, inpatient and community health nursing, pharmacy, and the QI department, plus a project coordinator from HealthTeamWorks [C.J.S.] and project evaluators from 16

UCD [S.P., D.S.M.]. This group, which first met in September 2005, guided the project, won support from stakeholders throughout the system, recruited key participants to design and implement interventions, and determined how to measure and report outcomes. In addition, three specialized work groups were active throughout the project: an IS group, a community health clinic group, and an inpatient group. An evaluation team collected and reported qualitative (interviews, field notes, meeting minutes) and quantitative (immunization rates) data to guide and assess the project as it was being implemented.

CLINICAL DECISION SUPPORT SYSTEM Unlike pediatric immunizations, for which routine recommendations are based only on age and vaccine history, adult immunization recommendations must also account for comorbid conditions and lifestyle factors. For this project, the IS team developed a CDSS for pneumococcal, influenza, and tetanus immunizations. A detailed algorithm was developed for each immunization on the basis of current US Centers for Disease Control and Prevention (CDC) recommendations. It included patient age, diagnoses (International Classification of Diseases, Ninth Revision [ICD-9] codes), and vaccine history. Diagnoses were obtained from an electronic problem list generated from billing data. The pneumococcal algorithm is displayed in Appendix 1 (available in online article). The CDSS searches data from the electronic patient clinical information database and a linked vaccine registry that stores immunization history and then generates either a message stating whether or not the immunization is recommended for a specific patient or a statement that the patient previously refused the immunization. The recommend message is supplemented with “yes/no” questions that query additional immunization indications that are not available in the patient medical record (for example, current smoking status). Developed as an on-demand system triggered by the user, the CDSS was intended to be used by members of the patient care team responsible for assessing immunization status and delivering immunizations. It was first available for use in outpatient community health centers and inpatient hospital units in January 2007. Pilot testing to integrate the CDSS into routine care was carried out in February–March 2007, followed by systemwide training and rollout through October 2007.

STANDING ORDER POLICY In March 2007 Denver Health approved a new provision in its standing order policy to authorize medical assistants and licensed nursing staff to carry out immunization delivery using

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

The Joint Commission Journal on Quality and Patient Safety the CDSS as a standing order in community health clinics. In the hospital, only registered nurses are authorized to utilize standing orders, and the CDSS replaced existing paper immunization standing orders in that setting.

STAFF EDUCATION Beginning in October 2005 in the hospital and in August 2006 in community health centers, clinicians and staff received periodic education on a range of topics such as influenza and pneumococcal immunization, immunization best practices (including screening for contraindications and documenting contraindications and refusals), clinical concerns (such as the low risk of pneumococcal revaccination in patients who lack documentation of previous immunization, and immunization considerations during pregnancy), and an orientation to the CDSS and standing order policy. Some portions of the education initiative began several months before the CDSS was implemented.

PILOT TESTING AND DISSEMINATION The CDSS and standing order policy were pilot tested during February and March 2007 using Plan-Do-Study-Act (PDSA) cycles in a family medicine community health clinic, an internal medicine community health clinic, and a general medical inpatient unit. In each setting, pilot testing lasted one month and was preceded with a review and mapping of the current immunization process to identify discrepancies and gaps, and standardize work flow. Rapid feedback of data on immunization rates for several patient subgroups (age 65 years or older, diabetes, or chronic obstructive pulmonary disease [COPD]) supplemented the implementation. After pilot testing, from April through October 2007 the CDSS and standing order policy were disseminated to the other community health clinics and the other adult inpatient medical units in the hospital. Leadership and coordination was provided by nurse managers in the community health centers and nurse educators in the hospital.

PROJECT COORDINATOR The HealthTeamWorks coordinator worked with the oversight group, all three special work groups, and the evaluation team throughout the project to schedule and facilitate meetings, conduct background research, communicate with Denver Health managers, draft immunization algorithms for the CDSS and review them with clinical experts, work on site with the pilot test clinics and hospital unit, and facilitate communication among all participants in the project.

DATA ANALYSES Immunization rates were analyzed using a mixed-effects model in which repeated measures over time were assumed to have an autoregressive [AR(1)] error structure. The marginal model was constructed using a simple linear spline that allowed the slope to change at 10 months; an additional indicator variable was added to allow a jump in the rate at 10 months for clinics receiving the late intervention.

Results Changes in pneumococcal immunization rates in the hospital and community health clinics during the course of the project are presented for three patient subgroups: adults 65 years of age or older, adults 18–64 years of age with COPD, and adults 18–64 years of age with diabetes (Figure 1, page 18).

COMMUNITY HEALTH CLINICS In the clinics a general trend toward increased immunization rates from July 2005 through March 2007 was already apparent. For the period before March 2007, it is unclear whether the QI initiative was responsible for the improvements, but after the adoption of the CDSS, immunization rates increased in the two pilot sites. After the CDSS was implemented in the other clinics between April and July 2007, rates increased by 10% overall.

HOSPITAL UNITS In the hospital units an overall trend toward improvement in pneumococcal rates before the introduction of the CDSS was also apparent, which was followed by an increase of 10% overall, after the nursing staff received training on pneumococcal and influenza immunizations and the new CDSS in February 2007. Statistical analyses (SAS version 9.1, SAS, Cary, North Carolina) of pneumococcal immunization rates in patients 65 years of age or older indicate a meaningful increase during the initial QI period when the CDSS was not in place (beta = 1.78, p = .05), but no additional significant increases were seen after this initial QI period (beta = 0.07, p = .93). A large initial increase in rates was observed in the remaining sites as the CDSS was widely adopted, though the change was not statistically significant (beta = 3.67, p = .20). Following the initial rate increase, there was a small statistically nonsignificant continued increase in rates among the rollout sites (beta = 0.67, p = 0.12).

RESPONSE TO A DECREASE IN IMMUNIZATION RATES One incidental finding that lends support for the potential of the CDSS to identify eligible patients for immunization is illustrated by the decrease in immunization rates noted in the

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

17

The Joint Commission Journal on Quality and Patient Safety Pneumococcal Vaccination Rates for Inpatients and Community Health Service Patients, Third Quarter 2005–Second Quarter 2008 Inpatient Pneumococcal Vaccination Rates

Community Health Service Pneumococcal Vaccination Rates

Figure 1. Pneumococcal vaccination rates are shown for inpatients (top) and community health service patients (bottom). The arrows on the figures identify key intervention points in the project. CDSS, clinical decision support system; COPD, chronic obstructive pulmonary disease; UTD, up-to-date on pneumococcal immunization; Q, quarter; 65+, 65 years of age or older; y.o., years old.

18

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

The Joint Commission Journal on Quality and Patient Safety COPD patient subgroup after January 2008 in the clinics and hospital. The IS department later determined that the ICD-9 code for COPD had inadvertently been dropped from the CDSS programming at that time and corrected the coding in 2009.

Discussion In this adult immunization improvement project in a large, integrated health care system, which was adapted from a successful pediatric immunization initiative, pneumococcal and influenza immunization rates started to increase in the pilot sites at the point at which the project coordinator started the QI work with Denver Health. The statistical model showed that the introduction of the CDSS did not change the trend of increasing rates over and above these initial QI efforts. However, had the initial improvement efforts not been supported later by the CDSS, rates might have decreased further toward the baseline after the project coordinator no longer facilitated the improvement process—as reflected in the decrease among patients with COPD that occurred when that ICD-9 code was inadvertently dropped from the CDSS. Rate increases were also observed in the non-pilot settings after the CDSS and standing order policies were implemented, lending support to the conclusion that improvements spread throughout the system. Compared with the usual paper standing order and reliance on clinical assessment, use of a computerized decision support tool—the CDSS—identified more patients with immunization indications. A standing order policy that authorized medical assistants in community health clinics to use the CDSS to guide immunization delivery took immunizations out of the hands of clinicians (except in the case of vaccine contraindications or patient refusals). This made the process of immunizing adults more efficient in busy primary care settings. Certain features of the CDSS developed for this project are particularly important for improved delivery of prevention services: It is computerized, it recommends a particular action, and it is available to the staff members who are most likely to carry out the service.14,15 In addition to the potential for immunizations received in an outpatient setting to prevent costly hospitalizations, there is also the advantage that reimbursement for outpatient immunization is higher than for the same immunization delivered in a hospital setting. The review of immunization practices and the delivery process that accompanied the introduction of the CDSS in the clinics and hospital provided opportunities to identify and correct oversights in documentation, retrain on immunization best practices, and improve the standardization of immunization work flow. The improvement process also highlighted the value

of providing ongoing feedback on immunization rates in a format that is useful to the clinic and hospital teams that provide patient care. Several other factors were important. Denver Health values improvement, innovation, and shared problem solving and there was clear leadership and management support for the project. Staff and financial resources were necessary to support the development of technological innovations. The existence of shared, interconnected information systems utilized throughout the institution was also essential. The project coordinator’s role in the planning, development, communication, education, and pilot testing was particularly important to keep the project on track because the initiative depended on collaboration and coordination among several departments of a large institution.

CHALLENGES AND SOLUTIONS During the course of the project it was necessary to address a variety of challenges, some unanticipated. The oversight and specialized work teams met regularly to monitor the project and identify, document, and address challenges and the approaches taken to resolve them. Table 1 (page 21) summarizes challenges and solutions, grouped into three categories: information systems, clinical, and sustainability. After implementation of the CDSS, the IS team received calls questioning the immunization recommendation in some patients (particularly pertaining to pneumococcal vaccine). In some cases, it was a matter of clinical knowledge about indications for immunization. In other cases, a review of the CDSS algorithm revealed that some of the ICD-9 codes included in the CDSS represented conditions that were not actually indications for immunization (for example, certain cardiac and pulmonary codes). The ICD-9 codes included in the CDSS were reviewed and revised in order to improve the specificity of the CDSS. The most important ongoing challenge has been inconsistent use of the on-demand CDSS as a routine part of patient care. Although the CDSS could have been designed to run automatically on opening a patient record, the oversight group chose to develop it as an on-demand system. Work flow varies among the hospital and clinics, and Denver Health encourages staff in each community health clinic to design and standardize work flow in a way that works best for the patient care process and for communication among team members. For example, in some clinics an immunization is given before a patient sees the provider, while in others the immunization is given at the end of the clinic visit. The CDSS is generally most useful on completion of an immunization assessment, whose timing varies. In some settings, the first person to open the patient record (for example, a front

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

19

The Joint Commission Journal on Quality and Patient Safety desk clerk who opens the patient record at registration) might not be the person responsible to act on the recommendation and provide the immunization. In that case, if the CDSS were automatic, the recommendation might not be reviewed and acted on by the right person, perhaps necessitating the introduction of another communication step into the process. In addition, because several other decision supports run automatically on opening the patient record, too many initial prompts could result in some recommendations being ignored. Whether automatic or on-demand, the CDSS must still be incorporated successfully into routine work flow and accompanied by consistent communication among all members of the practice team who have the potential to help improve immunization delivery. On inpatient units, in particular, nurses must complete several steps in the immunization process with the potential for inconsistency. The nurses complete an immunization assessment during patient admission; then order the vaccine(s) from the pharmacy; administer the vaccine(s); and then document the immunization(s), refusal(s), or contraindication(s) in three separate computerized systems. Some of these steps will be streamlined when the new electronic health record (EHR) is implemented (see page 22). In addition, inpatient units identified patient refusal of immunization as a common barrier, so there is a need for ongoing education of nurses on ways to encourage patients to accept immunization. Nurses on the different inpatient units have varying levels of familiarity and comfort with immunizations as a part of routine adult care.

PROGRESS SINCE THE QUALITY IMPROVEMENT PROJECT Since the end of the QI project in July 2008, Denver Health has continued to review and occasionally modify immunization work flow, perform chart audits (in the hospital) to troubleshoot missed opportunities, retrain on the CDSS and immunization best practices, and expand the CDSS to include other immunizations. A Denver Health adult immunization committee, some of whose members served on the project’s oversight team, has continued to meet regularly to monitor adult immunization progress, identify needs and opportunities (for example, by developing mechanisms to enable the hospital and clinics to routinely project vaccine needs on the basis of patterns of usage), and plan future initiatives. Tracking of Immunization Rates. Although funding was not available for continuing tracking of immunization rates as conducted during the project, Denver Health has continued to track and report new standard quarterly adult immunization rates to community health centers and hospital units, with some changes 20

introduced over time in the methods used for analysis. For example, more subgroups than diabetes and COPD are now included in the analysis of pneumococcal immunization in high-risk adults younger than 65 years of age; patients who died during hospitalization or checked themselves out of the hospital against medical advice are now excluded from analyses in the inpatient setting; and throughout the system, individuals who refused immunization are now counted as unimmunized, whereas they were previously counted as immunized if the refusal was documented. The month-to-month pneumococcal and influenza immunization rates vary across the community health centers, clinics, and hospital units, serve as a valuable tool for continued QI efforts, and emphasize the need for ongoing efforts to standardize and integrate the CDSS and immunization process into routine work flow. Modification and Spread of CDSS to Other Immunizations. Modifications have been made to the original CDSS for pneumococcal, influenza, and tetanus immunization, and new CDSSs have been developed for other immunizations. The ability to quickly make modifications has made the CDSS a powerful and adaptable tool. For example, in 2008, when the CDC changed adult influenza guidelines to include all patients,23 the CDSS was modified to base the recommendation only on age and documented vaccine history. In 2009, when the CDC added asthma, drug abuse, and smoking as indications for pneumococcal immunization,24 new ICD-9 codes were easily added to the CDSS, and a question was added to the recommendation message to identify current smoking. In the event of a national vaccine shortage, immunization recommendations in the CDSS could be modified to identify the highest-risk patients. A CDSS algorithm can also include payer source, which can help ensure that a patient receives the immunization in the optimal setting for payment for the service. One incentive to Denver Health for introducing the CDSS and standing order process into outpatient clinics was that reimbursement for immunizations delivered in the outpatient setting is higher than that in the inpatient setting. A CDSS can also be customized for different areas of the health care system to account for particular patient categories such as postpartum status, which is useful, for example, in the combined tetanus, diphtheria, and pertussis (Tdap) CDSS. A new CDSS for H1N1 influenza vaccine which was developed in fall 2009 included specific recommendations for both the nasal and injectable formulations of the vaccine. Previously, the inpatient Mom/Baby Unit was routinely providing measles, mumps, and rubella (MMR) vaccine using a paper standing order. A new MMR CDSS was developed in 2009 that includes

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

The Joint Commission Journal on Quality and Patient Safety Table 1. Three Types of Challenges and Their Solutions* Type of Challenge

Description

Solution

Information Systems (IS)

Technology infrastructure necessary to transfer immunization data to electronic patient record

Existing technical expertise and financial resources were reallocated for IS to develop the infrastructure.

IS

Balancing sensitivity and specificity of CDSS (for example, the pneumococcal algorithm initially included all ICD-9 codes for chronic lung disease)

Clinicians reported specific inappropriate immunization recommendations to IS; ICD-9 codes in the CDSS were evaluated and edited.

IS

Limited IS staff resources

Project time line was adjusted to accommodate competing priorities.

IS

Limitations of technology: running new immunization rules simultaneously along with existing rules unexpectedly overtaxed the system

IS team reconfigured the electronic systems to increase capacity.

IS

Cost of technology: development and long-term maintenance of systems

Management support and dedicated financial support were mobilized to develop and maintain systems.

Clinical

Most current inpatient hospital admission diagnoses not available to the CDSS

Requires nurse review of admitting diagnoses to assess for immunization indications

Clinical

Medication indications for immunization (for example, immunosuppressive medication) not recorded in the electronic record

Requires clinician consideration of additional indications for immunization

Clinical

Consent and screening process cannot be completed on patients with delirium or dementia.

Requires standardized protocol for addressing immunizations in these patients and regular review by nursing staff

Clinical

Inpatient process for ordering vaccine introduces delayed immunization administration.

Immunization administration was added to the scheduled medication administration system.

Clinical

Other indications for immunization (for example, resides in long term care facility, caretaker for young child) were not in the electronic record.

Requires additional “yes/no” questions administered by the nurse or medical assistant to ascertain these indications for immunization; scripting was developed and implemented.

Clinical

Ensuring that there is a standardized communication process between staff and clinicians about refusals/contraindications

Each site developed a process for communication and regular review of the process in staff meetings.

Clinical

Inaccurate ICD-9 codes in patient’s electronic problem list led to incorrect immunization recommendations.

Clinicians were asked to report inaccurate ICD-9 codes so that they could be removed from patient’s problem list.

Sustainability

Updates in vaccine indications (for example, smoking and asthma) were added as indications for pneumococcal immunization after the initial CDSS coding was completed.

Oversight of the clinical information in the CDSS was undertaken by the Denver Health Adult Immunization Committee.

Sustainability

Cost of immunizing all eligible patients

Pharmacy had to factor these costs into their budget.

Sustainability

Inconsistent use of the CDSS across patient care settings

Clinic and hospital unit teams must customize how the CDSS is used and coordinated with other applications such as computerized prescriber ordering.

Sustainability

Maintaining an adequate vaccine supply to meet increased demand

Ongoing communication with pharmacy and projection of vaccine supply needs

Sustainability

Maintaining staff knowledge of immunization best practices

Immunization training was included as part of newemployee orientation and in periodic staff retraining.

Sustainability

Maintaining improved immunization rates

Provide regular data feedback to clinic and hospital teams.

Sustainability

Navigating interdepartmental decision-making process

Communication, consultation with decision makers, good documentation of process; agency immunization committee oversight

Sustainability

Making the improvement unavoidable

Expectation was disseminated that CDSS would be used routinely at all clinic visits and on admission to the hospital.

* CDSS, clinical decision support system; ICD-9, International Classification of Diseases, Ninth Revision.

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

21

The Joint Commission Journal on Quality and Patient Safety a feature that queries electronic laboratory data to assess the presence or absence of rubella antibodies; if absent, immunization is recommended. This CDSS saves staff time because nurses no longer need to manually look up laboratory data. The MMR CDSS algorithm is shown in Appendix 2 (available in online article). A CDSS for genital human papillomavirus (HPV), developed in 2009, was the first to be tailored to a multidose vaccine. It calculates when the next dose in the series is due and provides an actual date (not a time frame) when the patient should return, a feature that saves staff time. The HPV CDSS also provides an alert message if the minimum interval between vaccine doses has not passed so that invalid doses of vaccine will not be administered, and it screens for an appropriate payer source. Discussions and planning are under way for the development of a CDSS for the zoster and meningococcal vaccines. The Mom/Baby Unit implemented the influenza CDSS and a new Tdap CDSS in 2009, and the inpatient psychiatry units and correctional care unit implemented the pneumococcal and influenza CDSS by 2010. Use of the immunization CDSSs has also spread to additional outpatient areas. By 2010 the Women’s Care Clinics adopted the CDSSs for HPV and Tdap. After becoming familiar with the technology, these clinics also started to use the influenza CDSS. Although the emergency department is not using the immunization CDSSs at this point, it is planning to begin to document immunizations in the Denver Health immunization registry to improve documentation of immunizations delivery. Making the Computerized Decision Support System Tool Automatic. Less-than-optimal immunization rates might reflect the limitations of an on-demand tool that is subject to individual variations in staff and provider practices and patient preferences and the need for ongoing monitoring to maintain and enhance improvements in immunization delivery. Use of the CDSS, an on-demand tool, relies on user initiation, which leaves open the possibility that it might not be used consistently at all visits. In 2012 Denver Health will implement an electronic health record (EHR), which should support improvements to the CDSS and its incorporation into routine work flow. In the new EHR, the CDSS will be able to run automatically on every patient encounter. The EHR will use process management software to enable triggers on events to initiate best-practice work flows. In addition, a dynamic problem list will permit nurses or physicians to enter diagnoses at the time of inpatient admission or at the first visit in an outpatient community health center and then make modifications to the list at follow-up encounters. This will improve the timeliness and accuracy of the documentation 22

of diagnoses, as compared with the current problem list, which is generated from billing data. Medication information will be available from Denver Health pharmacy data and from outside facilities through a medication reconciliation process. Some aspects of medication ordering and administration will also be simplified.

Conclusion This adult immunization improvement project illustrates how interdepartmental collaboration to build a strong infrastructure that incorporates computerized clinical decision support tools, standing orders, work-flow analysis and standardization, use of the most appropriate members of the patient care team to deliver services, and ongoing measurement of results can lead to sustainable improvements in immunization rates. The project may serve as a model for improved delivery of other preventive health services. J Support for the article described in this project was provided by The Colorado Trust. The authors acknowledge Diane Fairclough for statistical modeling of the data.

Online-Only Content

8

See the online version of this article for Appendix 1. Pneumococcal Algorithm Appendix 2. MMR Algorithm

Carolyn J. Swenson, MSPH, MSN, is Senior Project Manager, HealthTeamWorks, Lakewood, Colorado. Alicia Appel, MD, is Associate Professor of Medicine, University of Colorado Denver School of Medicine and Denver Health and Hospital Authority. Moira Sheehan, MAS, is Senior Application Analyst, eHealth Services; Anne Hammer, BA, BSN, RN, is Nursing Clinical Coordinator, Immunization Program; and Zita Fenner, BSN, RN, CNE, is Clinical Nurse Educator, Nursing Services, Denver Health and Hospital Authority. Stephanie Phibbs, MPH, is Senior Professional Research Assistant, University of Colorado Denver. Marjie Harbrecht, MD, is Chief Executive Officer, HealthTeamWorks. Deborah S. Main, PhD, is Professor, Health and Behavioral Sciences, University of Colorado Denver. Please address correspondence to Carolyn J. Swenson, [email protected].

References 1. US Department of Health & Human Services. Healthy People 2010 Goals and Objectives for Improving Vaccination Coverage. Accessed Nov 30, 2011. 0-www.ncbi.nlm.nih.gov.elis.tmu.edu.tw/books/NBK14084/#A8537. 2. Maciosek MV, et al. Methods for priority setting among clinical preventive services. Am J Prev Med. 2001;21(1):10–19. 3. Centers for Disease Control and Prevention: Interim Guidance for Use of 23Valent Pneumococcal Polysaccharide Vaccine During Novel Influenza A (H1N1) Outbreak. Jul 9, 2009. Accessed Oct 2, 2009 http://www.cdc.gov/h1n1flu/guidance/ppsv_h1n1.htm.

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

The Joint Commission Journal on Quality and Patient Safety 4. Centers for Disease Control and Prevention, National Center for Health Statistics. Early Release of Selected Estimates Based on Data from the January–June 2009 National Health Interview Survey. Accessed Jan. 12, 2010. http://www.cdc.gov/nchs/data/nhis/earlyrelease/200912_04.pdf/; http://www.cdc.gov/nchs/data/nhis/earlyrelease/200912_05.pdf. 5. Nowalk MP, Zimmerman RK, Feghali J. Missed opportunities for adult immunization in diverse primary care office settings. Vaccine. 2004 Sep 3;22(25–26):3457–3463. 6. Yarnall KS, et.al. Primary care: Is there enough time for prevention? Am J Public Health. 2003;93(4):635–641. 7. National Foundation for Infectious Diseases (NFID). Saving Lives: Integrating Vaccines for Adults into Routine Care. Bethesda, MD: NFID, 2008. 8. Gamble GR, Goldstein AO, Bearman RS. Implementing a standing order immunization policy: A minimalist intervention. J Am Board Fam Med. 2008;21(1):38–44. 9. Goebel LJ, Neitch SM, Mufson MA. Standing orders in an ambulatory setting increases influenza vaccine usage in older people. J Am Geriatr Soc. 2005;53(6):1008–1010. 10. Nowalk MP, et al. Increasing pneumococcal vaccination rates among hospitalized patients. Infect Control Hosp Epidemiol. 2003;24(7):526–531. 11. Bates DW, et al. Ten commandments for effective clinical decision support: Making the practice of evidence-based medicine a reality. J Am Med Inform Assoc. 2003;10(6):523–530. 12. Garg AX, et al. Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: A systematic review. JAMA. 2005 Mar 9;293(10):1223–1238. 13. Amarasingham R, et al. Clinical information technologies and inpatient outcomes: A multiple hospital study. Arch Intern Med. 2009 Jan 26;169(2):108–114. 14. Gerard MN, et al. Use of clinical decision support to increase influenza vaccination: Multi-year evolution of the system. J Am Med Inform Assoc. 2008;15(6):776–779.

15. Kawamoto K, et al. Improving clinical practice using clinical decision support systems: A systematic review of trials to identify features critical to success. BMJ. 330(7494)):765–772. Epub 2005 Mar 14. 16. Shortell SM, Bennett CL, Byck GR. Assessing the impact of continuous quality improvement on clinical practice: What it will take to accelerate progress. Milbank Q. 1998;76(4):593–624,510. 17. Bradley EH, et al. A qualitative study of increasing beta-blocker use after myocardial infarction: Why do some hospitals succeed? JAMA. 2001 May 23;285(20):2604–2611. 18. Bradley EH, et al. Quality improvement efforts and hospital performance: Rates of beta-blocker prescription after acute myocardial infarction. Med Care. 2005;43(3):282–292. 19. Berwick DM. Disseminating innovations in health care. JAMA. 2003 Apr 16;289(15):1969–1975. 20. Jones KL, et al. Improving adult immunization rates in primary care clinics. Nurs Econ. 2008;26(6):404–407. 21. Appel A, et al. Improving adult immunization delivery with policy changes and clinical support technology. Patient Safety &Quality in Healthcare. 2008. 5(5):32–36. 22. Melinkovich P, et al. Improving pediatric immunization rates in a safety-net delivery system. Jt Comm J Qual Patient Saf. 2007;33(4):205–210. 23. Fiore AE, et al. Prevention and control of influenza: Recommendations of the Advisory Committee on Immunization Practices (ACIP), 2008. MMWR Recomm Rep. 2008 Aug 8;57(RR-7):1–60. 2008 Jul 17 (Early Release). Accessed Nov 29, 2011. http://www.cdc.gov/mmwr/preview/ mmwrhtml/rr57e717a1 .htm. 24. Centers for Disease Control and Prevention. Recommended adult immunization schedule—United States, 2009. MMWR Recomm Rep. 2009 Jan 9;57(53). Accessed Nov 29, 2011. http://www.cdc.gov/mmwr/preview/ mmwrhtml/mm5753a6.htm.

How many people in your organization need The Joint Commission Journal on Quality and Patient Safety? Consider a site license. Having a Journal site license means anyone in your organization gets affordable, electronic delivery of all the great new Journal content directly to their computer every month. Plus they’ll have complete, searchable, electronic access to nearly 1,000 archived Journal articles dating back to 2000. Call 800-746-6578 today for more information on a site license for the Journal or any other Joint Commission Resources periodical.

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

23

The Joint Commission Journal on Quality and Patient Safety Online-Only Content

8 Appendix 1. Pneumococcal Algorithm* What is the patient’s age?

>=65 years

19-64 years

Current Medical History Indication (see list of ICD codes*)?

History of PPV vaccination (ever)?

Yes

No Yes

Time elapsed since PPV?

No

Recommend PPV: Proceed to patient screening History of PPV vaccination (ever)?

<5 years

>=5 years

Age at last PPV? Up-To-Date: Do not Recommend PPV <65

Recommend PPV: Proceed to patient screening

Yes

No

Up-To-Date: Do not Recommend PPV

Recommend PPV: Proceed to patient screening

>=65

Up-To-Date: Do not Recommend PPV

QUESTION Pneumovax is not recommended for this patient unless they are a smoker. Ask the patient if they smoke 10 or more cigarettes a day. If yes, the vaccine IS RECOMMENDED.

* ICD, International Classification of Diseases, Ninth Revision; PPV, pneumococcal polysaccharide vaccine.

AP1

Ask the patient the question listed below regarding smoking history

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

The Joint Commission Journal on Quality and Patient Safety Online-Only Content

8 Appendix 2. MMR Algorithm*

* MMR, measles, mumps, rubella.

January 2012 Volume 38 Number 1 Copyright 2012 © The Joint Commission

AP2