Value-Based HIV Care: a method to improve quality of HIV care.


 Background : To support our goal of providing optimal HIV care to our patients, we started applying the value-based health care principle to the HIV care that we offer in our HIV center, measuring relevant health outcomes and costs to allow continuous implementation of improvements (Value-Based HIV Care; VBHiC). Methods : In line with the principles of Michael Porter, our approach consisted of the following steps: 1) Organizing into integrated practice units / describing the HIV care path; 2) Defining an HIV outcome indicator set; 3) Building an enabling information technology platform; 4) Integrating care delivery across separate facilities; 5) Moving to bundled payments for care cycles and 6) Expanding excellent services and interventions for improvement across geographic boundaries. Results : The following set of 9 outcome indicators was developed: undetectable HIV load within the first year of care; quality of life within the first year of care; mortality within the first year of care; retention in care; therapy effectiveness; therapy tolerance; cardiovascular risk; quality of life for every subsequent year and overall annual mortality. These indicators, which were evaluated retrospectively, are shown in figures 1-5. Collection of the underlying data started in January 2016. The HIV care path was also integrated into the electronic file system. Creation of the ability to monitor outcome indicators at patient level, population level and process level allowed us to implement a quality cycle (plan-do-study-act). Conclusion : Our Value-Based HIV Care approach facilitated structured evaluation of parameters that are of value to the patient. It also boosted the quality of the HIV care that we provide and allowed us to increase the number of patients to whom we can offer high quality HIV care.


Methods
OLVG Hospital, a teaching hospital in Amsterdam's city center with 600 beds, cares for more than 500,000 patients per year, including approximately 3500 people living with HIV (PLHIV). We applied the six steps described by Porter to create a high value HIV care delivery system, as described below. We collected data on health outcome indicators and process indicators retrospectively from electronic patient files for the years 2005 through 2015, with prospective data collection starting in 2015.
Step 1) Organizing into integrated practice units / describing the HIV care path The HIV care path starts when the patient is referred to our HIV center. At each point in the care path, we described all necessary diagnostic and therapeutic steps in compliance with our formal standards and national guidelines 6 and developed clear descriptions of the roles played by the various actors in the care path, such as the patient, the nurse and the doctor's assistant. We labelled each visit in the care path with a specific code and attached a set of specific activities (orders) to each code.
These orders represent items such as laboratory tests and quality of life questionnaires, as well as items such as follow up visits. Where necessary, additional tests and visits can be manually added to the care path. Diagnostic tests and therapeutic activities relating to comorbidities were registered outside of the HIV care path. We defined two main HIV care paths: one for the first year of care and one for subsequent years. We also developed parallel care paths for specific, frequently occurring comorbidities such as hepatitis C infections, as well as for mental health problems such as anxiety and depression.
Step 2) Defining an HIV outcome indicator set The international consortium for health outcome measurement (ICHOM) issued a number of health outcome standard sets designed to transform health care systems worldwide based on standardised measurement and reporting of patient outcomes 7 .
These standard sets did not include a standard set for HIV health outcomes, however, so we decided to develop one ourselves. We began by searching the literature for terms such as HIV, monitoring, outcome, quality of care, morbidity, comorbidity, mortality, cause of death and admissions. We needed a set of outcomes that covered the full cycle of care for the patient. According to Porter, the outcomes that matter to patients for a particular medical condition fall into three tiers: Tier 1 is "Health status achieved or retained", Tier 2 is "Process of recovery" and Tier 3 is "Sustainability of health". We set up three panels of our own patients to perform indepth reviews of potential indicators in a focus group setting. In a working group with HIV health care professionals, a psychologist and a psychiatrist, we identified various quality of life domains based on evaluations of patient referrals for psychological problems over a ten-year period. We also assessed the potential indicators against the results of a quality of life and needs survey conducted by the Dutch HIV patient organization (HVN) of a representative sample of 468 PLHIV in the Netherlands 8 .
Ultimately, we selected the 9 outcome indicators with the highest prevalence in the population, the highest impact on the patient and the highest potential to be influenced by the patient and/or caregiver, or a combination of these factors. We Step 3: Building an enabling information technology platform The HIV care path, with all of its steps, the related codes and the underlying orders, was integrated into our electronic patient file system EPIC ® 9 in this step, which resulted in a detailed overview of our care process and the ability to extract discrete data at patient level, process level and population level, thus facilitating assessment of the HIV care that we offer our patients.

Step 4: Integrating care delivery across separate facilities
In this step, we described and formalized collaboration between the various health care specialisms and caregivers involved in HIV care, including specialists in the areas of pharmacology, medical microbiology, dermatology and psychiatry, as well as technicians in the hospital laboratory and doctors in general medical practices outside of the hospital.

Step 5: Moving to bundled payments for care cycles
In this step, we studied the selected indicators in close consultation with a key national health insurer and used the results to develop a new bundled payment contract to cover patient care. The contract is quality-driven and provides for a predetermined payment per patient per year.
Step 6: Expanding services across geographic boundaries.
This step, which is ongoing, involves the development of formalized methods, guidelines, processes and best practices, based on measurable and verifiable benchmarks and showcased in the application of Value-Based HIV Care in our HIV center at OLVG Hospital, in preparation for a roll out to other HIV centers throughout the Netherlands and elsewhere.

Results
Step 1: Organizing into integrated practice units / describing the HIV care path Table 1 is a schematic overview of key moments in the patients care path. After referral, an HIV nurse will see the patient within 3 working days for an intake with a duration of one hour (appointment code "0.0"). In cases involving urgent medical problems, the patient will also see an HIV specialist at the same moment. Then, blood and urine analysis are performed, as well as a chest X-ray, and the patient is asked to complete his or her first PROM, which consists of 62 questions. Where possible, the patient is started on combination antiretroviral therapy (cART) during this first intake. When cART is initiated, the patient is given detailed information on the prescribed therapy by an HIV nurse. The HIV nurse, who is assigned to the patient and acts as the patient's contact person for any questions relating to HIV and the prescribed therapy, maintains regular contact with the patient after the therapy is initiated (via the telephone and face-to-face appointments in our HIV center). A faceto-face appointment is scheduled for all patients 4 weeks after cART is initiated (special code for appointment 4 weeks after start or switch cART). In cases in which the patient does not immediately see an HIV doctor due to urgent medical problems, an appointment is scheduled for a 30-minute consultation with an HIV doctor within 2 weeks of the initial intake (code "0.1"). During this consultation, the HIV doctor reviews the patient's medical condition, including the results of cART if therapy was prescribed during the intake (code 0.0). In cases in which the patient has not yet been given therapy, cART is initiated during this consultation. Patients are seen again by the HIV doctor three months later (code "0.4"; 10 minutes) and subsequently every 4 months until the end of the first year (code "0.8" (10 minutes) and code "1.0" (20 minutes)). Code "1.0" marks the end of the patient's first year of care and represents a 20-minute consultation with the HIV doctor. During this consultation, the patient receives the results his or her annual check-up, which includes lipid profile and blood glucose readings and urine analysis results, as well as a discussion of the patient's PROMs. Patients are asked to complete PROMs before every consultation with the HIV doctor in the first year of care.
After the first year, the patient enters the chronic phase of follow-up, with visits every six months (code "1.6" (10 minutes) and code "2.0" (20 minutes)). The second of these two visits (code "2.0"), similarly to the code "1.0" visit at the end of the first year of treatment, includes a review of the patient's PROMs and lab results. During the chronic phase, in the absence of comorbidities or other complications, patients are seen alternately by an HIV nurse (1.6 NS (nurse specialist); 20 minutes) and the HIV doctor (2.0 MS (medical specialist). Where this delegation of tasks to the HIV nurse is unfeasible, the patient is seen twice a year by the doctor (1.6 MS and 2.0 MS). In cases involving complications and comorbidities, the patient will be seen more frequently. In these cases, appointment codes that differ from codes used in the HIV care path are used to distinguish the diagnostic procedures and health care activities performed to handle complications and comorbidities from those that fall within the HIV care path.

Step 2: Defining an HIV outcome indicator set
Our search of the literature 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 11 yielded 38 potential outcome indicators (Appendix A). The nine that we selected to form our HIV outcome indicator set are shown in Table 2. Step 3: Building an enabling information technology platform The electronic patient file system EPIC ® enables us to monitor our HIV care at patient level, process level and population level. Parameters necessary for indicator surveillance are registered using a special HIV form that is built into EPIC ® and that also makes discrete data storage possible. All VBHiC parameters are displayed on a single screen so that the patient and caregiver can see how the patient is performing on each of the parameters, as well as what data may be missing for that patient. All VBHiC data and processes are closely monitored by a data analyst for accuracy and compliance. The data from the system are used for internal evaluation and quality improvement purposes. Each month, a working group of doctors and nurses from our team meets to assess one of the outcome indicators. During these working group meetings, the effects of earlier interventions are evaluated and, if necessary, additional interventions are implemented to continuously improve care through a plan-do-check-act (PCA) cycle.

Step 4: Integrating care delivery across separate facilities
The HIV care path includes optional referrals to other facilities. Optional referrals include referrals for anal and cervical dysplasia screening, for instance, as well as referrals to a psychologist or psychiatrist in cases involving raised anxiety/depression scores (HADS >15). In the future, we plan to integrate the roles of other caregivers in the HIV care path into our VBHiC system. Among the caregivers we plan to integrate into the system are general practitioners, pharmacists, medical microbiologists and dermatologists.
Step 5: Moving to bundled payments for care cycles A three-year value-based health care contract with a key national health insurance provider that provides for bundled payments for HIV care went into effect in January 2018. The contract, which was signed in December 2017 and includes quality parameters from our VBHiC program, contains strict definitions for the care that is covered by the HIV care bundle and for the care that is covered via other provisions.
Two quality indicators were included for the first year, based on the levels of these indicators in 2016: cardiovascular risk evaluation and cascade of care as a composite indicator of mortality, retention in care and therapy effectiveness. Future plans will include additional quality parameters such as HIV mental health/quality of life, which will help to improve projections of HIV-related health care expenditures outside of the hospital. The contract also includes provisions for cost monitoring and follow-up in time. Figure 5 shows a decline in total laboratory costs and outpatient department (OPD) consultation costs despite growth in the patient population.
Step 6: Expanding services across geographic boundaries.
Our primary goal in developing our VBHiC program was to improve the care that we

Discussion
Incorporating value-based HIV care into our daily clinical practice forced us to redefine our care process based on outcome parameters with an added value for our patients. We defined our key outcome parameters in close collaboration with our Dutch patient organization to optimize added value. With a detailed HIV care path (step 1), we are able to focus greater attention on important moments in the HIV care path, thus heightening the visibility of activities within the care path and clarifying expectations, for both the HIV caregiver and the patient. Patients receive the support they need to arrive fully prepared for their consultations in the outpatient clinic and the structured approach helps us to incorporate easily missed but essential aspects Therapy tolerance (indicator 5) was a relatively difficult parameter to define. Toxicity can be defined objectively in biochemical or radiological terms (kidney function, lipids, bone mineral density), but the term 'switch due to toxicity' is more subjective and dependent on the viewpoints and attitudes of the doctor and the patient over time.
We were unable to find a validated questionnaire for scoring medication tolerance or intolerance. In practice, we expect the structural inclusion of therapy tolerance in our questionnaire to increase patient satisfaction levels.
Evaluation of cardiovascular risk (indicator 7) is now an integral part of the annual consultation. By discussing the 10-year cardiovascular risk with patients and focusing on decreasing cardiovascular risk to less than 10% for patients 11 we expect use of statins and medication to reduce hypertension to increase in the short term, but we also hope to persuade more patients to stop smoking as time progresses. Hopefully, our efforts in this area will result in a decline in cardiovascular events and a corresponding decrease in overall cardiovascular costs.
Routine measurement of quality of life using PROMs (indicators 2 and 8) was the most innovative outcome indicator in our VBHiC approach. From the moment they were introduced, the PROMs turned out to be a good point of departure for structured discussions of physical and mental health with our patients. Most patients appreciate the holistic approach that the PROMs represent. Approximately 60% of patients returned their questionnaire in the first year. Of course, logistical problems such as language barriers, privacy concerns and incorrect email addresses need to be overcome with any questionnaire, and it takes time to implement and familiarize patients with the process. Patients who are more difficult to reach may even simply refuse to complete PROMs. We recently set up a project to analyze why certain patients fail to complete PROMs and to develop other methods to evaluate quality of life for these patients.
Our collaboration with our hospital's psychiatry department is an example of integrating care deliveries across separate facilities (step 4). In the past, we lacked objective criteria for referring patients to the outpatient clinic of psychiatry department. In the VBHiC program, we now use a hospital anxiety depression score as a screening tool and indicator for referrals to the psychiatry department. In addition, attending multidisciplinary meetings and dual consultations with our HIV nurses has resulted in more active involvement in HIV care on the part of the hospital's psychologists and psychiatrists.
Effective communication, an efficient organization and a computer-based care path are required to implement and sustain value-based HIV care. Our HIV care process is built into EPIC ® . Other HIV centers will need to build their own electronic patient file systems. We have a VBHiC working group for operational issues and a steering group to monitor progress and facilitate future planning and embedding within the hospital organization. The indicator working groups report their initiatives and results every 3 months. Our VBHiC approach attracted the attention of one of the Netherlands' largest health insurer and resulted in a 3-year bundled payment contract (step 5) that we expect to benefit both our patients and the insurer. Other HIV centers in the Netherlands have begun to implement VBHiC based on our outcome indicators (step 6). In the future, we will all benefit from the opportunities for benchmarking and learning from best practices that an ever-widening implementation of VBHiC will bring.
Value-based HIV care closed a gap in the HIV care path, creating a structured system for the evaluation of parameters that are valuable to our patients. It offers us objective, measurable quality indicators for HIV care and has helped us to offer a growing number of patients excellent HIV care without significant increases in the number of caregivers and without significant increases in the overall cost of care.

Declarations: Ethics approval and consent to participate
The ecthical committee of pour hospital ("ACWO OLVG") approved collection of all used data to improve our care and to submit these data for publication

Consent for publication
The manuscript does not contain any individual person`s data

Availability of data and materials
Data supporting the results reported in the article are stored in the " data warehouse" of OLVG hospital and can be obtained if requested.

Competing interests
There are no competing interests for any of the authors

Funding
No funding was obtained for this manuscript

Authors` contributions
All authors have contributed substantially to the conception and design of the work and have approved to the submission of the current manuscript. Furthermore, they have agreed to be personally accountable for the authors` own contributions.