How Compliance Measures, Behavior Modification, and Continuous Quality Improvement Led to Routine HIV Screening in an Emergency Department in Brooklyn, New York

Abstract

Objective

New York State adopted a new HIV testing law in 2010 requiring medical providers to offer an HIV test to all eligible patients aged 13–64 years during emergency room or ambulatory care visits. Since then, Wyckoff Heights Medical Center (WHMC) in Brooklyn, New York, began implementing routine HIV screening organization-wide using a compliance, behavior-modification, and continuous quality-improvement process.

Methods

WHMC first implemented HIV screening in the emergency department (ED) and evaluated progress with the following monthly indicators: HIV tests offered, HIV tests accepted, HIV tests ordered (starting in December 2013), HIV tests administered, positive HIV tests, and linkage to HIV care. Compliance with the delivery of HIV testing was determined by the proportion of patients who, after accepting a test, received one.

Results

During August 2013 through July 2014, of 57,852 eligible patients seen in the WHMC ED, a total of 31,423 (54.3%) were offered an HIV test. Of those, 8,229 (26.2%) patients accepted a test. Of those, 6,114 (74.3%) underwent a test. A total of 26 of the 6,114 patients tested (0.4%) had a positive test, and 24 of the 26 HIV-positive patients were linked to HIV medical care. By July 2014, the monthly proportion of patients offered a test was 62%; the proportion of those offered a test who had a test ordered was 98%, and the proportion of those with a test ordered who were tested was 81%. Testing compliance increased substantially at the WHMC ED, from 77% in December 2013 to >98% in July 2014.

Conclusion

Using compliance-monitoring, behavior-modification, and continuous quality-improvement processes produced substantial increases in offers and HIV test completion. WHMC is replicating this approach across departments, and other hospitals implementing routine HIV screening programs should consider this approach as well.

In 2013, the U.S. Preventive Services Task Force found convincing evidence that earlier identification and treatment of human immunodeficiency virus (HIV) infection were associated with a markedly reduced risk for progression to clinical illness and premature death.¹ The Task Force also found that earlier treatment could substantially decrease the risk of HIV transmission to uninfected partners or from mother to child at birth. These findings formed the basis for the Task Force's Grade A recommendation for routine HIV screening, meaning the Task Force found that there is high certainty that the net benefit of HIV screening is substantial and that practices should offer this service. Grade A recommended preventive health services are covered by insurers. As such, the recommendation paved the way for increased implementation of HIV screening in U.S. health-care systems. This recommendation followed a 2006 Centers for Disease Control and Prevention (CDC) recommendation that also called for routine, opt-out HIV screening in all health-care settings in which newly diagnosed HIV prevalence rates were at least 0.1%.²

In compliance with CDC recommendations, New York State adopted a new HIV testing law in September 2010 that requires medical providers to routinely offer an HIV test to patients aged 13–64 years.³ If test results are positive, medical providers are required to provide results in person, conduct posttest counseling, and help the patient make an appointment to begin HIV medical care. If negative, test results can be returned by the provider (or representative) in person, by phone, via mail, or electronically. In April 2014, the state modified the law to allow for oral consent, which was consistent with consent for other routine health-care services and which further streamlined HIV testing.

New York City (NYC) has the largest HIV epidemic of any city in the United States, ranking first in the number of people living with HIV infection in 2010 and in the number of people newly diagnosed with HIV infection in 2011.⁴ As in much of the rest of the country, important disparities in HIV infection exist. Individuals newly diagnosed with HIV in 2012 in NYC were primarily male, black or Hispanic, young, men reporting sex with men, or people living in impoverished areas. Compared with people living in low-poverty neighborhoods, those living in high-poverty neighborhoods also experienced more HIV-related deaths and had poorer long-term survival. Even in relatively low-poverty NYC neighborhoods, black or Hispanic people with HIV had poorer survival rates.⁵

Wyckoff Heights Medical Center (WHMC) in Bushwick, Brooklyn, is in a community that has, historically, been at the epicenter of the HIV epidemic in NYC.⁵ WHMC has provided HIV medical treatment and other services to people living with HIV infection, but most people who presented for care were already in advanced stages of illness (Unpublished data. New York City Department of Health and Mental Hygiene, 2012). Beginning in August 2013, WHMC began to integrate routine HIV screening across the institution to comply with the new state testing law requiring that an HIV test be offered to ensure that those living with HIV are identified more promptly to avoid poor health outcomes.

This article describes lessons learned during WHMC's implementation of routine HIV screening. We discuss our experience in terms of compliance monitoring and continuous quality improvement. We further extend the definition of compliance—going beyond an offer of HIV testing—to include the actual administration of the HIV test because we noticed a substantial gap between those who accepted the offer and those who were actually tested. We hope that sharing our experiences will not only help our own community, but also help achieve the goals of the U.S. National HIV/AIDS Strategy: to get people living with HIV diagnosed and into high-quality medical care, and to reduce HIV-related health disparities.⁶

METHODS

Wyckoff Heights Medical Center

WHMC, established in 1889 as a not-for-profit teaching hospital, has 324 beds serving Northern Brooklyn and Queens. The hospital provides a full range of inpatient and ambulatory care, including emergency services. In 2013, WHMC saw 155,000 patients: 36% Hispanic, 28% Caucasian, 27% African American, 5% Asian/Pacific Islander, and 4% other race (Unpublished data. WHMC electronic medical records, 2013).

Development of the WHMC routine HIV screening program

In August 2013, WHMC began an organizational change process to integrate routine HIV screening into all areas of the institution, beginning with the emergency department (ED). WHMC approached the project as a quality-improvement and compliance initiative, and began by conducting a root-cause analysis to identify the specific areas to target for improvement. One key area of improvement was ED staff member perception, motivation, and attitudes toward change in HIV testing procedures. WHMC addressed these concerns and the change process by using Bullock and Batten's integrative model for planned change, which is divided into four phases: exploration, planning, action, and integration.⁷

In the exploration phase, WHMC created a team dedicated to establishing training, providing solutions, supporting staff members, reinforcing policy, and providing feedback on routine HIV screening procedures. The team identified ways to improve issues with electronic medical records (EMRs), recruited provider champions who were staff members of the ED system and would support the program, and formed a committee to review compliance measures and recommend improvements.

In the planning phase, the team conducted interviews with key ED staff members who reported that medical providers might resist offering the HIV test to all patients. Most medical providers felt that HIV testing should continue to be delivered through a dedicated counselor model because patients would require counseling and intervention at the time the test was offered. Staff members also believed that they did not have time to provide the HIV test given their other duties in a busy ED, nor were they equipped to administer it. Another potential barrier was a high patient throughput: the ED had 60,000 patient visits in 2012. This number was key for medical providers who believed that integrating HIV testing in the ED would inevitably cause further delays and unnecessarily burden an already stressed system. To address this issue, the team analyzed the patient process flow from arrival to discharge. EMR prompts to offer tests were developed to alert staff members to make the offer and track the patients who accepted it. At the 30-day point after implementing the new HIV testing process and periodically thereafter, patient wait time was evaluated and found to not have increased as a result of the offer and administration of an HIV test. The medical providers found collecting specimens for the HIV test could be naturally incorporated into the patient's care for those whose blood was being drawn. For those patients who were not getting blood drawn, the fast-track area administered the blood draw during triage, which allowed the laboratory to process the sample more readily and have the results available once the patient had seen the doctor.

Furthermore, providers felt ill-equipped and reluctant to manage patients' emotional reactions to receiving a preliminary positive HIV test result. To relieve this anxiety, providers were assured that the HIV prevention program and Frontlines of Communities in the United States (FOCUS) program teams⁸ would continue to assist with delivery of HIV-positive results. We found that with this knowledge in hand, the medical providers demonstrated a newfound confidence and successfully delivered positive test results to their patients, often on their own. The planning phase also involved staff member in-service trainings, supervision sessions, and executive committee meeting presentations on the new routine HIV screening procedures.

The daily continuous quality-improvement reviews allowed for a thorough evaluation of missed opportunities (i.e., those patients who accepted the offer of an HIV test but were not tested). In three cases, the patients came to the ED with flulike symptoms, fatigue, and diarrhea. Because no HIV test was performed, the patients' symptoms were treated and they were released only to return again several weeks later with the same symptoms. During the subsequent visit, the patients received the HIV test they had accepted at registration, and all three patients received a preliminary HIV-positive test result. Sharing the results of these case studies with our medical providers and executive leadership was a catalyst for positive change. After discussions, we realized how important it was for all involved to be able to put a face to the missed tests. Recognizing that missed tests equated to individual lives, fewer providers were opposed to offering and administering the HIV tests.

In the action phase, WHMC implemented the changes for the ED identified during the planning phase and began tracking test-level data of the HIV testing continuum: offers, acceptances, tests ordered, tests conducted, test results provided, results delivered, and (for those who were positive) linkages to HIV medical care. Patients excluded were those in trauma or <13 years of age, unless requested by the parent. The offer of an HIV test was an opt-in process that took place at registration, but transitioned to an opt-out process as the NYC HIV testing law was revised. The team had support from the WHMC hospital executive team and several HIV testing champions from the committee. The committee was also instrumental in developing appropriate quality improvements and monitoring their impact on testing outcomes. This continuous quality improvement required detailed and near-real-time reporting of HIV testing outcomes at the individual provider level. The outcomes were key to ensuring policy compliance during the action phase.

During the integration phase, the team sponsored an event highlighting and acknowledging staff member efforts to effectively integrate HIV screening into the ED. This event occurred just six months after the launch of the initiative. To support continued integration, results and outcomes from the changes were communicated with ongoing reports throughout the organization. Training and education for employees remained an ongoing priority.

WHMC routine HIV screening process

WHMC patients aged 13–64 years who were not already diagnosed with HIV and had no documented HIV test in the past six months were eligible for a test offer. The EMR system determined this eligibility automatically and prompted staff members to make the offer during registration and as part of the ED visit's triage steps. Once the patient accepted the HIV test from the triage nurse, an automated system that tracks pending doctor's orders on registered ED patients was triggered, allowing the EMR system to register the patient's acceptance of the offer and add the patient's name to the accepted list, which served to remind the practitioner to draw blood and order the HIV test. In addition to the list, the medical team could easily see a red “y” (for “yes”) next to the patient's name, which signified that the patient accepted an offer for testing. The patient returned to the waiting room to await registration. After registration, the nurse tracker was illuminated, which prompted the nurse to order the HIV test. The nurse would call the patient in to retrieve a specimen before the visit with the medical provider.

Our EMR system was not equipped to automatically generate laboratory orders. Furthermore, once the test was ordered, the specimen had to be drawn and sent to the laboratory. As the team found during its hourly real-time compliance monitoring of test orders, these steps in the process were identified as two opportunities where testing could be missed. If those who accepted a test had not had one ordered but were still in the ED, a team member contacted the ED provider to determine why the test was not ordered. If the EMR did not acknowledge the ordered test as being received in the onsite laboratory within 30 minutes of the order, a team member would contact the ED provider or laboratory to determine why the ordered test was not received.

The HIV serologic test used was the Abbott ARCHITECT^® Ag/Ab Combo test (Abbott Laboratories, Inc., Abbott Park, Illinois). The ED provider retrieved HIV test results from the EMR and gave them to the patient prior to discharge, or the patient returned at a later date for a copy of the negative result. For positive HIV test results, the laboratory immediately contacted the team to assist the ED provider in delivering the positive result to the patient and making the linkage to HIV medical care.

Program evaluation methods

To evaluate the routine HIV screening program's implementation, we examined the following HIV testing continuum indicators by month: offers of an HIV test, acceptance of an HIV test, orders of an HIV test (starting in December 2013), HIV tests conducted, and positive HIV tests. All of this information was abstracted from the EMR for unique patients. Patients who visited the ED multiple times during the year could be counted up to two times in the data because they could be tested twice in a year through WHMC's managed care contract. Compliance for delivery of HIV testing to those who accepted was illustrated using two different approaches: (1) monthly proportion of individuals who had been tested vs. proportions of those who did not have a test ordered or who had a test ordered but did not have a test conducted, and (2) weekly number of missed tests among those who accepted testing and were examined by an ED unit (e.g., fast track, adult, and pediatric). We also described overall demographic characteristics of the patients who were tested for HIV and those who tested positive for HIV. For those who tested positive, we described the linkage to HIV medical care (defined as having attended the first HIV medical care appointment), and the CD4+ count and HIV viral load immediately after their HIV diagnosis.

RESULTS

From August 2013 through July 2014, WHMC saw 57,852 eligible patients, offered 31,423 (54.3%) patients an HIV test, had 8,229 (26.2%) of those patients accept an HIV test, and conducted 6,114 (74.3%) HIV tests. The majority of patients who were tested for HIV were female, aged 23–40 years, and Hispanic (Table). A total of 26 (0.4%) patients were newly identified as HIV positive, of whom 24 (92.3%) were linked to HIV medical care and two died before linkage to care could occur. The majority of patients who had a positive HIV test were male, 31–40 or ≥51 years of age, and non-Hispanic black (Table). Of the 18 (69.2%) patients with CD4+ count data, the mean CD4+ count was 279 cells/cubic millimeter (mm³) (95% confidence interval [CI] 188, 371). Six patients had a CD4+ count <200 mm³, six patients had a CD4+ count of 200–349 cells/mm³, four patients had a CD4+ count of 350–500 cells/mm³, and two patients had a CD4+ count >500 cells/mm³. Among 16 patients with HIV viral load data, the mean log-10 viral load was 2.2 (95% CI 1.4, 3.1).

Table.

Number of HIV tests administered and HIV-positive test results, by patient characteristics, Wyckoff Heights Medical Center, Brooklyn, New York, 2013–2014

^aPercentages may not total to 100 due to rounding.

^bPercent positive is the row percent (i.e., the number of positive tests divided by the total number of tests for patients with each characteristic).

HIV = human immunodeficiency virus

From August 2013 through July 2014, increases occurred in the number and proportion of eligible patients who were offered tests (from 937 of 4,684 [20.0%] to 3,167 of 5,108 [62.0%]), the proportion of patients who accepted testing who had a test ordered (from 721 of 933 [77.3%] to 718 of 730 [98.4%]), and the proportion of those who had a test ordered who were tested (from 164 of 391 [41.9%] to 703 of 871 [80.7%]) (Figure 1). Testing compliance initially varied by ED unit, with the greatest number of missed tests occurring in the fast track (i.e., the unit in the ED where patients with minor injuries receive prompt care). The pediatric unit had the fewest missed tests. The fasttrack and adult units exhibited substantial improvement in the number of missed tests on a monthly basis (Figure 2). The number of HIV tests relative to missed tests also substantially improved, from 164 tests taken and 226 tests missed in August 2013 to 703 tests taken and 77 tests missed in July 2014.

Figure 1.

HIV testing program outcomes, by month, Wyckoff Heights Medical Center emergency department, Brooklyn, New York, August 2013–July 2014

^aGray bars are the number of HIV tests conducted each month.

HIV = human immunodeficiency virus

Figure 2.

Number of HIV tests accepted but not conducted (i.e., missed tests), monthly per unit, Wyckoff Heights Medical Center emergency department, Brooklyn, New York, December 2013–July 2014

^aFast track refers to the unit in the emergency department where patients with minor injuries receive prompt care.

HIV = human immunodeficiency virus

LESSONS LEARNED

After implementation of the routine HIV screening program at WHMC, the offering and administration of HIV tests increased substantially. Some of the most encouraging feedback we received from our medical providers was that this success has helped rekindle our team's desire for early identification, early treatment, and a healthier population. Sharing case studies of specific patients who were previously missed from the accepted offer listing and were subsequently identified as HIV positive showed the medical team the impact of not testing individuals who accepted a test offer. We shared these findings, and the legal and fiscal ramifications of missed testing on our institution, with our executive team.

Our willingness to share our system's deficiencies and subject ourselves to scrutiny was beneficial. We received feedback from executive leadership and got the internal support we needed to move forward with systemic improvements. It also allowed us to take corrective action against those who failed to adhere to the policies that had been developed to not only support the New York State HIV testing law, but also to ensure continuity in a standard of excellence in our patient care. This support was critical to successfully incorporating systemic, sustainable change.

The importance of compliance

CDC guidance on routine HIV screening in health-care settings was published in 2006,² and the New York State law requiring health-care providers to offer HIV testing and to simplify informed consent was enacted in 2010.⁹ To meet our obligations for routine HIV testing, we needed to recognize not only that our systems needed improvement, but also that change is always approached from a compliance standpoint in health care. Health-care providers respond to compliance guidelines, such as those used by the Joint Commission on Accreditation of Healthcare Organizations,⁸ to monitor adherence to clinical quality standards.¹⁰ As such, we turned to a simple checklist we use as a guide to implementing any compliance initiatives to ensure all areas were being addressed. The checklist has six requirements, listed here along with how we specifically addressed each requirement:

• Code of conduct: During trainings, we reminded staff members of the Code of Ethics for Nurses.

• Written policies and procedures: We updated our HIV policies to allow documentation of oral consent for HIV testing to be based on acceptance of the test as noted in the EMR. We also obtained our forms committee's approval to house the newly created test result letters in our EMR system.

• Educational and training programs for staff members and physicians: The medical providers had thought they were ill-equipped to communicate with their patients about the HIV test offer or the test results. Our trainings included candid discussions that allowed for open dialogue and one-on-one coaching, eventually empowering staff members with the knowledge and understanding necessary to accomplish their new tasks.

• Effective communication mechanisms: We -developed a protocol of daily real-time communication. Each day, we shared the previous day's compliance (i.e., how many patients had accepted the test offer and were actually tested). In addition, adding the electronic indicator on the nursing tracker provided a visual communication that a patient had accepted the offer and needed to be tested. We attribute much of our success to this system upgrade.

• Auditing and monitoring systems: We developed audit and monitoring processes to enforce individual accountability. The medical providers were aware of our presence; we not only reviewed the number of HIV tests accepted after the test offer and the number of tests administered, but we also reviewed the providers' notes. We encouraged them to use their notes to communicate any outliers that we were unable to see at the front end; that is, we asked that they give us the tools we needed to accurately report their compliance on a daily basis. This cooperation provided yet another layer of communication without us having to ask why a person was not tested and strengthened their documentation of services rendered.

• Appropriate disciplinary and corrective action measures: Even with our best efforts to foster a culture of patient safety, we still saw patterns of accepted HIV tests being missed repeatedly by nurses and providers, despite reminders from the monitoring staff members. In these instances, we used corrective training sessions to encourage change and adherence to policy. These sessions included discussions with staff members to identify barriers they might have encountered, sharing data and patient outcomes that the monitoring team was looking at reviewing with the clinical staff members, and re-reviewing the HIV testing policy. We then monitored the system again. If non-desired outcomes persisted, a supervisor could choose to take administrative measures (i.e., counseling sessions to help the employee and achieve the outcome). No punitive actions were necessary to get nurses and providers to adhere to policies.

Challenges ahead

The dramatic increase in HIV testing can quickly deplete an institution's budget and laboratory's resources if we fail to ensure all systems are working congruently. In fact, a preliminary review of billing revealed that we had not been coding tests appropriately to receive all reimbursement. We are working to ensure that HIV testing can be funded through proper billing and reimbursement because higher levels of testing in the ED requires an additional nurse technician (due to the increased volume of phlebotomy).

We are also expanding our initiatives to our outpatient clinics by conducting both monitoring and modification of current workflow processes, technology enhancement, and medical provider trainings. Our baseline data revealed that, although the HIV test offer and acceptance rates increased in outpatient clinics, we were not testing those who had agreed to be tested. Primarily using the nursing team in the ED was most beneficial. However, due to differences in workflow, reduced acuity of ailments, and the need to support a physician-driven model of care for all tests and treatment, including HIV and hepatitis C, we have restructured and educated our outpatient medical team.

Mandating compliance has drawbacks. We must review all issues cautiously, paying attention to the cost and ramifications of our actions. However, this initiative allowed for a systemic evaluation and revealed opportunities to revamp our billing practices, ultimately leading to the generation of funding to sustain this compliance initiative. The workflow changes have been supported and sustained with current staffing. The majority of our medical team has accepted the execution of the HIV test as part of their normal workflow and has communicated with us directly that they are simply doing their job. At six months, noncompliance was <5%; with the support of ED leadership, coaching, and counseling, we are beginning to see those numbers decrease further. Additionally, our next objective is to increase the number of people who accept our test offer. We believe that our transition from an opt-in HIV test offer methodology to an opt-out approach will result in greater penetration of our eligible population. We are beginning a new campaign to increase the number of people who say “yes” to knowing their status.

CONCLUSION

Our compliance initiatives are in their infancy, but our approach—meticulous in nature and demanding deliberate analysis and monitoring—is working. Compliance, behavior modification, and continuous quality improvement all influence workplace behavior, motivation, and performance, leading, in our case, to an assurance that all patients who accept an HIV test offer at WHMC will, indeed, be tested. We will no longer accept non-testing as our standard, and we will continue to address any barriers through real-time accountability and education.