PHM Glossary: P

Pareto Principle or the 80/20 rule

An application of the Pareto principle to health care utilization is the frequent observation that a small percentage of any population (often 10% or 20%) accounts for a large percentage of that population’s resource utilization, or claims. Often the percentage of the population’s resources accounted for by the high-utilizing fraction is the complement of that group’s percentage of the total population, hence, the name 80/20 rule.


The chart shows the distribution of members by net paid claims band for members with chronic conditions, non-chronic members, and certain catastrophic members of a health plan over the period of 1 year. Overall, 81% of members have net paid claims less than $1,000 in a year.

Distribution of Members within each Subpopulation
Group < $1,000 $1,000 - $9,999 $10,000 - $49,999 $50,000 - $99,999 $100,000+ TOTAL
Chronic 52.04% 39.07% 8.42% 0.41% 0.06% 100%
Non-Chronic 82.33% 16.62% 1.01% 0.03% 0.00% 100%
Excluded 83.91% 13.33% 2.20% 0.32% 0.24% 100%
TOTAL 81.00% 17.35% 1.52% 0.09% 0.04% 100%

Members who have annual costs less than $1,000 (81% of membership) account for just 19.5% of claims. The distribution of members and claims is, however, very different for the chronic and non-chronic populations. Only 52% of chronic members have costs less than $1,000 and these costs amount to only 6.6% of chronic member cost. 82% of non-chronic members have costs less than $1,000, but because non-chronic costs are lower than chronic costs, this group accounts for more than 25% of all non-chronic member costs. Thus we conclude that:

  1. The Pareto principle is true for the population as a whole and for chronic and non-chronic subpopulations.
  2. The often-repeated form of the Pareto principle, while it is a useful shorthand way of describing the principle, may or may not be accurate, depending on the specific population being considered. In fact the “rule” may be misleading for some populations.

A second example is a 2003 study performed by the Wisconsin Hospital Association; the study also noted that those with chronic illnesses (5% of the overall population) used 33% of resources.

Distribution of Members within each Subpopulation
Group < $1,000 $1,000 - $9,999 $10,000 - $49,999 $50,000 - $99,999 $100,000+ TOTAL
Chronic 6.61% 37.94% 46.21% 6.77% 2.47% 100%
Non-Chronic 25.37% 52.43% 19.30% 2.12% 0.78% 100%
Excluded 9.44% 24.06% 29.98% 12.70% 23.82% 100%
TOTAL 19.46% 44.81% 25.46% 4.85% 5.43% 100%


Duncan, I. “Part 2: Actuarial Issues in Care Management Interventions.” Society of Actuaries. 23 Feb. 2006

Patterson, D. “Outpatient Services, in Allies and Adversaries: The Impact of Managed Care on Mental Health Services.” Washington, DC: American Psychiatric Publishing (1994).

“Wisconsin’s Healthcare Environment.” Wisconsin Hospital Association. 23 Feb. 2006


A participant in a population health management program is an eligible health plan member who has had at least one contact with a population health management program and has not opted out of the population health management program. For opt-in programs, an individual who has opted in is a participant. For opt-out programs, individuals are considered participants unless they take action to be excluded.


A patient is an individual who receives care or services. Other related terms include member, client, resident, customer, or health care consumer.


Joint Commission on Accreditation of Healthcare Organizations. Disease-Specific Care Certification Manual (2002).

Patient Activation

Patient activation refers to an individual’s knowledge, skills and confidence to play a role in their health care. The Patient Activation Measure (PAM) is a validated instrument that measures the attributes of an activated consumer. PAM assesses four stages of activation: (1) belief that the patient’s role in their own care is important, (2) having the confidence and knowledge necessary to take action, (3) taking action to maintain and improve one’s health, and (4) maintaining health behaviors while under stress.

Level of activation is influenced by age, education, culture, literacy, health literacy, as well as personal preferences. By using PAM to identify a patient’s stage of activation, physicians can individualize their care plans. PAM can also be used to assess how well such interventions work.


Hibbard JH, Stockard J, Mahoney ER, Tusler M. Development of the Patient Activation Measure (PAM): Conceptualizing and Measuring Activation in Patients and Consumers. 2004; 39(4): 1005-1026.

Patient Advocate

A patient advocate is an individual who helps a client and the family work with others who have an effect on that client’s health. Ideally, the patient advocate should advocate for both the client and the payer to facilitate positive outcomes.


A patient advocate helps resolve issues at the service-delivery level, the benefits-administration level, and the policy-making level with issues about health care, medical bills, access to services, and job discrimination related to a patient’s medical condition. However, in conflict the needs of the client must be the first priority. Patient advocacy groups may also try to raise public awareness about important issues related to specific diseases (e.g., cancer, AIDS, or renal disease), such as the need for disease-specific support services, education, and research.

The patient advocate will work to ensure that a comprehensive assessment identifying client-specific needs occurs; that options for necessary services are provided; and that clients are provided with access to resources for meeting their individual needs. Additionally, patient advocates focus on:

  • Establishing effective relationships with all health care stakeholders (client/family, payer, physician, other health care providers, and other relevant parties);
  • Promoting client well-being;
  • Promoting client/family informed decision making and independence;
  • Encouraging and supporting the move toward self-care;
  • Educating and assisting in the facilitation of access to necessary and appropriate health care services; and
  • Working for systems improvement.


Case Management Society of America. Standards of Practice for Case Management (2002).

Commission for Case Manager Certification. Code of Professional Conduct for Case Managers (2004).

National Association of Social Workers. Specialty Certification in Case Management (2004).

The Health Encyclopedia. 2004. 23 Feb. 2006

Patient-Centered Medical Home (PCMH)

The Patient Centered Medical Home (PCMH) is an approach to providing comprehensive primary care for children, youth and adults. The PCMH is a health care setting that facilitates partnerships between individual patients, and their personal physicians, and when appropriate, the patient’s family.


The American Academy of Pediatrics (AAP) introduced the medical home concept in 1967, initially referring to a central location for archiving a child’s medical record. In its 2002 policy statement, the AAP expanded the medical home concept to include these operational characteristics: accessible, continuous, comprehensive, family-centered, coordinated, compassionate, and culturally effective care.

The American Academy of Family Physicians (AAFP) and the American College of Physicians (ACP) have since developed their own models for improving patient care called the “medical home” (AAFP, 2004) or “advanced medical home” (ACP, 2006).


The AAP, AAFP, ACP, and AOA, representing approximately 333,000 physicians, have developed the following joint principles to describe the characteristics of the PCMH.

  • Personal physician - each patient has an ongoing relationship with a personal physician trained to provide first contact, continuous and comprehensive care.
  • Physician directed medical practice – the personal physician leads a team of individuals at the practice level who collectively take responsibility for the ongoing care of patients.
  • Whole person orientation – the personal physician is responsible for providing for all the patient’s health care needs or taking responsibility for appropriately arranging care with other qualified professionals. This includes care for all stages of life: acute care, chronic care, preventive services, and end of life care.
  • Care is coordinated and/or integrated across all elements of the complex health care system (e.g., subspecialty care, hospitals, home health agencies, nursing homes) and the patient’s community (e.g., family, public and private community-based services). Care is facilitated by registries, information technology, health information exchange and other means to assure that patients get the indicated care when and where they need and want it in a culturally and linguistically appropriate manner.
  • Quality and safety are hallmarks of the medical home:
    • Practices advocate for their patients to support the attainment of optimal, patient-centered outcomes that are defined by a care planning process driven by a compassionate, robust partnership between physicians, patients, and the patient’s family.
    • Evidence-based medicine and clinical decision-support tools guide decision making
    • Physicians in the practice accept accountability for continuous quality improvement through voluntary engagement in performance measurement and improvement.
    • Patients actively participate in decision-making andfeedback is sought to ensure patients’ expectations are being met
    • Information technology is utilized appropriately to support optimal patient care, performance measurement, patient education, and enhanced communication
    • Practices go through a voluntary recognition process by an appropriate non-governmental entity to demonstrate that they have the capabilities to provide patient centered services consistent with the medical home model.
    • Patients and families participate in quality improvement activities at the practice level.
  • Enhanced access to care is available through systems such as open scheduling, expanded hours and new options for communication between patients, their personal physician, and practice staff.
  • Payment appropriately recognizes the added value provided to patients who have a patient-centered medical home. The payment structure should be based on the following framework:
    • It should reflect the value of physician and non-physician staff patient-centered care management work that falls outside of the face-to-face visit.
    • It should pay for services associated with coordination of care both within a given practice and between consultants, ancillary providers, and community resources.
    • It should support adoption and use of health information technology for quality improvement;
    • It should support provision of enhanced communication access such as secure e-mail and telephone consultation;
    • It should recognize the value of physician work associated with remote monitoring of clinical data using technology.
    • It should allow for separate fee-for-service payments for face-to-face visits. (Payments for care management services that fall outside of the face-to-face visit, as described above, should not result in a reduction in the payments for face-to-face visits).
    • It should recognize case mix differences in the patient population being treated within the practice.
    • It should allow physicians to share in savings from reduced hospitalizations associated with physician-guided care management in the office setting.
    • It should allow for additional payments for achieving measurable and continuous quality improvements.


Add cite:


Patient Portal

Health care related online applications that allow patients to interact and communicate with their health care providers, such as physicians and hospitals. Typically, portal services are available on the Internet at all hours of the day. Some patient portals applications exist as stand-alone web sites and sell their services to health care providers. Other portal applications are integrated into the existing web site of the health care provider. Still others are modules added onto an existing electronic medical record system. All of these share the ability of the patient to interact with their medical information via the Internet.

Patient Safety

Patient safety is the prevention of harm to patients. Patient safety efforts aim to reduce errors of commission or omission.


Patient safety, according to the Institute of Medicine (IOM), is the prevention of harm to patients and can be addressed at multiple levels. Population health management programs are uniquely positioned to perform four functions important to patient safety: communication, care coordination, analysis and education.

Population health programs can incorporate patient-centric systems and technology through, for example, data mining and routine claims analyses both at the aggregate and patient-specific level. Population health management has been proactive in medication safety by implementing electronic medical databases and has laid the groundwork for the next generation of medical safety with patient partnering at the center of the strategy.

Several patient safety definitions are currently in use:

  • Centers for Medicare and Medicaid Services (CMS) – “Patient safety is the condition or act of freeing patients from the risk of harm, injury, or loss inherent from their interaction with the health care delivery system independent of the risk of harm, injury, or loss imposed from their particular disease process.”  
  • URAC – “An adverse event result(ing) in unintended harm to the patient by an act of commission or omission rather than by the underlying disease or condition of the patient.”
  • Centers for Disease Control and Prevention (CDC) – “Patient safety, broadly defined, should incorporate preventive services. Since a lapse in safe practices increases the risk-of-harm for a patient, then the failure to provide a necessary preventive health service can properly be considered a safety lapse because it increases the risk of illness or other harm.”
  • National Patient Safety Foundation – “Patient safety is the prevention of health care errors, and the elimination or mitigation of patient injury caused by health care errors.”

In addition, The Joint Commission has established national patient safety goals (NPSGs) to help accredited organizations address specific areas of concern in regards to patient safety. The NPSGs are health care setting specific and are updated annually.


These are defined as unintended health care outcomes that are the result of a defect in the delivery of care to an individual. These errors may be made or caused by any member of the health care team in any health care setting. Patient safety efforts address two types of errors committed by the health care system—commission and omission.


  • Not using evidence-based medicine (e.g., applying outmoded therapies)
  • Inappropriate medication
  • Error in dose or use of medications (e.g., drug-drug interactions)
  • Lack of patient-centered medication management (e.g., not obtaining patient preference or considering cognitive or functional status for dosing and/or route etc. to ensure adherence)


  • Failure to use indicated tests (e.g., kidney damage and/or uncontrolled incontinence with bladder dysfunction /retention)
  • Incomplete follow-up of therapy
  • Lack of interdisciplinary care management (e.g., lack of coordination and collaboration) to close gaps in care
  • Lack of timely referrals for needed disciplines or specialties (e.g., podiatry services for neuropathy screening and/or appropriate orthotics for patients with diabetes or advanced practice/specialty physicians or nurses)
  • Lack of adequate depression and/or anxiety assessment and treatment
  • Lack of comprehensive, patient-centered education regarding self-care management (i.e., environmental, psychosocial. and readiness for change assessments are omitted) and coping techniques for expected disease trajector
  • Lack of timely assessment for adherence problems (e.g., follow up telehealth, telephonic and/or interactive video or other monitoring, Internet-based and/or mailed questionnaires or home visits)
  • Lack of appropriate referral to community-based services (e.g., nonprofit organizations with comprehensive community-based services and patient support groups)
  • Lack of timely ordering of evidence-based equipment (e.g., support surface to prevent a pressure ulcer for chair-bound and/or bed-bound patients)
  • Lack of comprehensive medication inventory (prescribed by providers and over-the-counter medications) and/or assessment for actual adherence issues
  • Lack of timely indicated risk assessment for factors such as cognitive, falls, or skin (using reliable and valid tools per evidence-based guidelines)


  • State of the science interdisciplinary, evidence-based protocols and clinician tools
  • Comprehensive patient and caregiver education regarding disease trajectory; strategies to manage symptoms, facilitate adherence, and prevent adverse events
  • Patient-centered strategies: Empowerment of patients in effective and efficient self-care, and maximization of self-care efficacy by aggressively addressing symptoms and/or co-morbidities affecting self-care adherence
  • Aggressive, timely, interdisciplinary care management: Reduction of avoidable adverse events, including rehospitalization and emergency room use.


Collection of comprehensive patient-specific data and information to support identification of quality of care and patient safety issues is a key component of population health management. These data can be aggregated with other available information such as claims data. Data mining and predictive risk modeling provide aggregate data to boost performance improvement, increasing disease-specific population safety. Population health management programs:

  • Use innovative systems and processes to systematically identify patients at risk of patient safety and quality-of-care issues;
  • Use systems and processes for interdisciplinary interventions—communicating patient safety efforts and identifying gaps in care to treating physicians and/or members; and
  • Reduce the need for residential health care placement by addressing environmental, psychosocial, functional, and symptom management issues.


Screening process for engagement – Telehealth (telephonic, interactive video, etc.) multidimensional population health management interventions ascertain changes in functional status (physical and/or cognitive) or changes in objective or subjective patient signs and symptoms. Timely screening identifies problems and promotes early intervention, before disease exacerbations occur.

Empowerment process for post enrollment – Empowerment of patients maximizes their self-care efforts. Patient education can address patient self-administration errors (e.g., missed doses of anticoagulant post open heart valve surgery etc.), subtherapeutic dose (not titrated to therapeutic level), toxic dose (for weight, age, etc. that’s not personalized or patient-centered) and missing medications indicated by best practice (evidence-based) guidelines.

For example, a patient in a population health management telemonitoring program has two episodes of feeling ill following ingestion of a heavy meal. A nurse case manager (NCM) received blood pressure, pulse, respirations pulse oxymetry, and weight remotely at the central station. Via telelephonic assessment, the NCM determines that the patient has been nonadherent with his low-sodium diet in both episodes.

The patient was motivated (ready for change) to prevent further rehospitalizations but admitted that occasional cheating or dietary indiscretion is likely to occur. The patient was a good candidate for new crucial self-care techniques. He was taught to monitor his daily weight, manage diuretic dosing, and call his physician to notify him before any rehospitalization. Thus, the patient is empowered in effective self-care, and there were no rehospitalizations for more than one year post-disease management intervention. The patient is no longer a high utilizer due to timely medication adjustments. The patient is coached through the first episode but makes the calls to the physician himself for the extra diuretic. The nurse communicates to the physician the patient’s learning status, readiness for change, etc.

Educational process for personalized knowledge and adherence efforts—Population health management reduces drug errors and/or reduces avoidable side effects, as well as ensuring therapeutic effect via telephonic, interactive interventions. In-home health management assessments yield increased accuracy of medication inventory and reported medication adherence. Promotion of evidence-based medicine adds to this effort.

Lack of appropriate self-administered and/or caregiver medication administration instruction can contribute to medication error: patients taking medications inappropriately, which interferes with appropriate action or predisposes the patient to avoidable, potentially dangerous side effects. Population health management comprehensive and follow-up instruction in medication self-care, periodic return demonstration, etc., improves medication administration safety at home.

Population health management incorporates technology at the systems level – Innovative use of new technology (i.e., telehealth), risk modeling (aggregate individual and population-specific data), and interdisciplinary approaches to timely interventions increase access to the right treatment at the right time for the right patient (improves care continuity and coordination).


Berwick, D. “A User’s Manual for the IOM’s ‘Quality Chasm’ Report.” Health Affairs 21(3)(2002): 80-90.

Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academies Press. 23 Feb. 2006

Data Standards for Patient Safety. 2006. Institute of Medicine of the National Academies. 23 Feb. 2006

Dixon, R. “Should Prevention Be a Part of a Patient Safety Initiative?” Presented at URAC Patient Safety and Medical Management Leadership Meeting, Washington, DC (2004).

Greenberg, L. URAC. Presentation at URAC Patient Safety and Medical Management Leadership Meeting, Washington, DC (2004).

Hunt, D. “Medicare Patient Safety Monitoring Systems.” Presented at: URAC Patient Safety and Medical Management Leadership Meeting, Washington, DC (2004).

NPSF Home Page. 2005. National Patient Safety Foundation. 23 Feb. 2006

“Reducing Medical Errors Requires National Computerized Information Systems; Data Standards are Crucial to Improving Patient Safety.” 20 Nov. 2003. The National Academies. 23 Feb. 2006

Patient Satisfaction

Patient satisfaction is the patient’s perception of the quality and value of the services and/or products received.


Patient satisfaction is a way of describing the degree to which an individual patient or health plan member regards the health care service or product. It is a measure of the patient’s perception of the services, rather than an objective evaluation of the services themselves. Patient satisfaction may also apply to whether the patient perceives the manner in which care is delivered by the provider as useful, effective, or beneficial.

Typically, patient satisfaction is measured via a survey and considered to have been achieved when the patient’s perception of the quality of care and services received in a health care setting has been positive, satisfying, and/or meets his/her expectations. The patient satisfaction survey may be conducted via telephone, mail, in person, or online. These surveys may cover care received during a specific office visit, hospital stay, telephone or e-mail encounter, or during an episode of care covering several encounters.

In general terms, patient expectations that are assessed in patient satisfaction surveys include:

  • Obtaining good treatment in a comfortable, caring, safe environment and delivered in a calm and reassuring way;
  • Receiving information to make choices, to feel confident and stay in control;
  • Being involved in health decisions to the extent desired;
  • Receiving good continuity of care; and
  • Being treated with honesty, respect, and dignity.


Survey scales are an important part of monitoring satisfaction. Availability of negative responses and outreach to nonrespondents strengthen the value of results. In addition, the use of an appropriate Likert scale and sampling methods such as surveying nonparticipating eligible members will increase the robustness of the survey results.


Paddock, L., et al. “Development and Validation of a Questionnaire to Evaluate Patient Satisfaction with Diabetes Disease Management.” Diabetes Care 23 (7) (2000): p 951-956.

Patient Treatment Plan

The patient treatment plan is a road map for guiding the multidisciplinary care team involved with a specific patient’s care to provide the most appropriate treatment to ensure the optimal outcome. The patient treatment plan is a living document that needs to be fluid and changeable in response to patient status and problem resolution.


The treatment plan is the result of thorough initial and ongoing assessments that generate a comprehensive plan that includes a problem(s) or need(s) list strategies/interventions to address identified problems or needs; goals with defined time frames for achievement, and anticipated responses, such as problem resolution, prevention or delay of complications, optimized quality of life, or comfort within the review period. All goals must be specific, measurable, and realistic. Likewise all stated interventions must be measurable and realistic. Patient treatment plans may also include available resources, medical review triggering, contractual requirements, and the desires/motivations of the client/family.


Case Management Society of America. Standards of Practice for Case Management. Little Rock, AR. (2002).

Sox, H. “What is a Care Plan?” Care Plans. 15 June 2005


The use of financial incentives to encourage and reward providers for the delivery of evidence-based practices associated with desired health outcomes.

These are also known as Pay for Excellence or Pay for Quality programs. They essentially all provide additional payments for focusing on quality outcomes; some pay additional incentives for administrative and service improvements.


Traditionally, physicians and other providers have been paid on a fee for service basis, independent of the outcome of those services. The Institute of Medicine (IOM) has recognized the need to transform physician payment. In its landmark publication Crossing the Quality Chasm, the following recommendations regarding physician payment were made:

  • Fair payment should be given for good clinical management for the types of patients seen. Clinicians should neither gain nor lose financially for caring for sicker patients or those with more complicated conditions;
  • Providers should have the opportunity to share in the benefits of quality improvement;
  • Consumers and purchasers should have the opportunity to recognize quality differences in health care and direct decisions accordingly;
  • Financial incentives should align with implementation of care processes based on best practices and the achievement of better patient outcomes and; and
  • Payment methods should reduce fragmentation of care.

A number of payers have implemented methodologies that increase incremental payments. These systems of reimbursement are often called “pay-for-performance” (“PFP” or “P4P”), because payers use financial incentives to affect targeted services and/or outcomes. Many of these programs include non-quality related metrics, such as metrics based on cost and administrative efficiencies which may be unrelated to the quality of care and services provided to consumers.


American Healthways/Johns Hopkins.“Outcomes-Based Compensation: Pay-for-Performance Design Principles” (2005).

Casalino, L. et al. “External Incentives, Information Technology, and Organized processes to Improve Health Care Quality for patients with Chronic Diseases.” Journal of American Medical Association. 289(4) 2003. 433-441.

Institute of Medicine. (2001). Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press, pages 183-184.

Per Member Per Month (PMPM)

Per member per month refers to the cost or revenue a health insurance plan receives from each plan member for a month. It includes revenue, expenses, or services use. In the context of lifestyle management or disease management programs, this refers to the price or revenue spread across a specified population (e.g., employees, employees/spouses or all health plan members).


Washington State Office of the Insurance Commissioner. A Consumer’s Insurance Glossary.

Performance-based Contract

A written agreement between the population health management organization and the client that puts some or all fees at risk to accomplish specified outcomes. Those outcomes may include participation rates, lower health care costs, satisfaction, clinical and financial outcomes, (including return on investment), and lower employee absenteeism.

If the population health management organization does not meet the specified performance levels, a portion of their fees may be refunded to the client. Alternatively, the agreement may be written so that if the performance levels are exceeded an additional payment is made to the disease management organization. The fees put at risk can be wholly financial or clinically or a combination of both (e.g. 100% clinical or 50% financial and 50% clinical).


See conformance to treatment and care plans.

Personal health record (PHR)

An electronic application through which individuals can maintain and manage their health information (and that of others for whom they are authorized) in a private, secure and confidential environment.

Pharmacy benefit manager (PBM)

A PBM is an organization that manages prescription drug benefits for self-insured clients, state employers, Medicaid and fully insured plans. Possible administrative services provided by the PBM can include contracting with pharmacy networks; establishing payment levels; negotiating rebate arrangements; developing and managing formularies and prior authorization programs; maintaining patient compliance programs; performing drug utilization review; and operating chronic care management programs. Many also operate or include access to mail order pharmacies. PBMs have a key role in the Medicare drug program.

Through a variety of strategies, PBMs offer clients ways to improve patient safety and health outcomes and to mitigate cost.


Nash DB, Reifsnyder J, Fabius RJ, Pracilio VP. Glossary of Population Health: Creating a Culture of Wellness. Sudbury, MA: Jones & Bartlett Learning; 2011.

URAC. Frequently Asked Questions about URAC’s Pharmacy Benefit Management Accreditation.


Population is a term used to refer to the aggregate or totality of all individuals who conform to a set of specifications.


General: The total number of diabetic members enrolled in Health Plan A.
Specific: The total number of diabetic members enrolled in Health Plan A above the age of 18 and without evidence of end-stage renal disease.


Total diabetic members in Health Plan A
Total diabetic members enrolled in Health Plan A continuously during the measurement period


Total diabetic members in Health Plan A.
Total diabetic members enrolled in Health Plan A at any time during the measurement period


Population is a group of persons with common qualities or characteristics. In statistical measurement, a sample or subset may be studied to make inferences about the population. In chronic care management, a population is all members of a defined group (health plan, insured group, employee group, etc.) that meet specific criteria that define them as eligible for a specific chronic care management program or intervention(s).

A population is simply an operational definition that describes the group of members included in the chronic care management program, pilot study, or evaluation. For example, definition of criteria for inclusion in a population could be based on certain chronic conditions and/or enrollment criteria.

The concept of continuous and dynamic populations is illustrated by the following chart:

Number of Periods   Exposure Duration
Both Complete Partial
Either Continuos Dynamic

A continuous population requires complete eligibility and enrollment in both baseline and intervention periods; a dynamic population may comprise members who are present in one or another of the periods, or who have partial exposure during these periods.


Health plan/group healthcare benefit eligible population: Eligible members (i.e., covered or insured members) for benefit coverage in the health plan or for an employer sponsored plan. Rules for eligibility coverage are typically well-defined in health plan and/or employer sponsored benefit plan contracts.

Program Eligible Population: Individuals meeting benefit coverage eligibility defined above and plan eligibility requirements for wellness or care management programs. Rules for program eligibility are typically well-defined in health plan and/or employer sponsored benefit plan coverage descriptions.

Identified Population for specific programs: Program eligible members who are identified as meeting wellness or care management program identification criteria.

Targeted Population: Subset of identified population who meet requirements for program qualifications and are targeted for wellness or care management program intervention.

Enrolled Population: Individuals enrolled in a wellness or care management program. Types of enrollment processes include:

  • Opt-in – Enrolled: Members consenting to participate in the program
  • Opt-out – Enrolled: Targeted population considered enrolled, unless action taken to disenroll.

Engaged Population:

  • Passive engagement strategies do not require active member consent or acknowledgement, program participation, and there are typically no defined action requirements on the part of the participant.
  • Active engagement strategies do require active member consent or acknowledgement of program participation, and defined action on the part of the participant.
  • The Initially Engaged Population is a subset of enrolled members who are working or have worked directly with a nurse or health coach in a chronic care management or health improvement program within the reporting period.
    • Members are interacting with the health professional in reference to their health improvement plan with “bidirectional interaction” meaning an exchange between the health professional and the member in both directions.
    • A participant is considered initially engaged if she has completed a clinical and lifestyle behavioral risk assessment that includes a mutually agreed upon plan of care with goals and at least one follow-up coaching discussion within 90 days.
    • Only real-time human interaction is included in this definition of initial engagement, regardless of the venue used for discussion.


Care Continuum Alliance. (2010). Outcomes guidelines report, vol. 5. Washington, DC: Care Continuum Alliance.

Care Continuum Alliance. Disease Management Program Evaluation Guide (2004).

Population Health Management

Population health management is an approach that aims to improve the health status of the entire population through coordination of care across the continuum of health in order to improve behavioral/lifestyle, clinical and financial outcomes.


This approach reflects a shift in thinking about how health is defined. A population health approach applied within chronic care management recognizes that health is a capacity or resource rather than a conditional state. Population health encompasses the realization that a range of physical, environmental, and socioeconomic factors (influencers) contribute to health. By successfully managing health influencers, population health endeavors to affect the complete physical, mental, and social well-being of a population.

The primary focus in population health management is to equip individual members within the population and those who influence and set policy with the necessary tools to make appropriate choices and decisions about their health and medical care, thereby achieving/ maintaining optimum health and reducing unnecessary medical expenses for the population.

A key component of success is adapting the environment along with providing education and support to facilitate behavioral change and make change more sustainable. Population health management approaches may include culture assessments/audits; policy changes; comprehensive needs assessments that assess potential/actual physical, social, psychological, economic, and environmental needs; proactive health promotion programs that increase awareness of the health risks associated with certain personal behaviors and lifestyles; patient-centric health management goals; and self-management interventions aimed at influencing the targeted population to make behavioral changes.


McAlearney, A. Population Health Management in Theory and Practice. JAT Press (2002).

Savage, E., Blair, J. and Fottler, M. Advances in Health Care Management: Vol (3). Chicago: Health Administration Press (2003).


Power is the ability of a study to statistically demonstrate an association among variables, if one exists.

The statistical power of a study is the probability that the study will detect that one group deviates from another if a difference truly exists between the groups. This is also called a deviation from the null hypothesis, where the null hypothesis states that two groups are the same. In population health management, there is interest in two treatments that differ from one another in cost and effectiveness.

The percentage of power tells the likelihood of being able to detect that difference if one exists. In epidemiological studies, seek a power of at least 80%, so that there is at least an 80% probability of detecting the difference between the two groups if one exists. Power is affected by several factors, including the sample size of the study, the actual magnitude of the difference (in means or proportions) that one is trying to detect (or the magnitude of the odds ratio in a case-control study), the variance, and the prevalence of the exposure in the controls (in a case-control study).

Technically, power = (1 – Probability of Type II error) = (1 – β), where Type II error (or β) is the probability of not rejecting the null hypothesis given that the null hypothesis is false. In other words, Type II error is the probability of finding that there is no difference between two groups when, in fact, there is one.


Last, J. A Dictionary of Epidemiology, 4th ed.Oxford: Oxford University Press (2001).

Spiegel, M., et al. Schaum’s Outline of Probability and Statistics. 2nd ed. New York: McGraw-Hill (2000).

Predictive Modeling

Predictive modeling is the process of forecasting future health care expenditures, resource utilization, or adverse clinical events (such as inpatient admits) based on differences in individuals’ health statuses. An important objective of predictive modeling is the allocation of health management resources to patients who are predicted to incur high health care expenses (in the absence of some kind of care intervention).

With the predictive modeling technique, individual patients are assigned a risk score based on their likelihood of using health services in the near future. In addition, predictive modeling techniques can be used to identify patients for control and intervention groups for population health management program evaluation.


Classification and regression trees are analytic procedures for predicting the values of a continuous response variable (e.g., health care cost) or categorical response variable (e.g., the top 20% most expensive cost group: yes or no) from continuous or categorical predictors. When the dependent or response variable of interest is categorical in nature, the technique is referred to as classification trees. If the response variable of interest is continuous in nature, the method is referred to as regression trees.

In the most general terms, the purpose of the analyses via tree-building algorithms is to determine a set of if-then logical (split) conditions that permit accurate prediction or classification of cases. For classification problems, the goal is generally to find a tree where the terminal tree nodes are relatively pure—i.e., contain observations that (almost) all belong to the same category or class; for regression tree problems, node purity is usually defined in terms of the sums-of-squares deviation within each node.

Advantages of classification and regression trees methods include the following. In most cases, the interpretation of results summarized in a tree is very simple. This simplicity is useful not only for purposes of rapid classification of new observations (it is much easier to evaluate just one or two logical conditions than to compute classification scores for each possible group or predicted value, based on all predictors and using possibly some complex nonlinear model equations) but can also often yield a much simpler model for explaining why observations are classified or predicted in a particular manner. For example, when analyzing prediction problems, it is much easier to present a few simple if-then statements to users than elaborate equations.

Tree methods are nonparametric and nonlinear. The final results of using tree methods for classification or regression can be summarized in a series of (usually few) logical if-then conditions (tree nodes). There is no implicit assumption that the underlying relationships between the predictor variables and the dependent variable are linear—follow some specific nonlinear link function, or are even monotonic in nature.

For example, some continuous outcome variable of interest could be positively related to a variable prior claims cost if the prior claims cost is less than some certain amount, but negatively related if it is more than that amount (i.e., the tree could reveal multiple splits based on the same variable prior claims cost, revealing such a nonmonotonic relationship between the variables). Thus, tree methods are particularly well-suited for data mining tasks, where there is often little a priori knowledge nor is there any coherent set of theories or predictions regarding which variables are related and how. In those types of data analyses, tree methods can often reveal simple relationships between just a few variables that could have easily gone unnoticed using other analytic techniques.


A concurrent (retrospective) predictive model uses information available from one time period to project medical claim costs for that same period. It is intuitive, in that concurrent models have greater explanatory power than prospective models, since more recent information is used for predictors. Generally, a concurrent model can be used for profiling or making retrospective adjustments to payments.


A hybrid model predicts a disease-related outcome (such as adverse clinical events, hospitalizations, or costs) by using several modeling methodologies in cooperation or competition. Hybrid modeling is generally used to predict outcomes for a specific data set. The model it produces is not intended for use on multiple, unrelated data sets. Therefore, hybrid models may be best predictors for the data on which the model was developed but do not generalize well to other sets.

Like all predictive models, a hybrid model attempts to predict who will acquire a disease, a disease-related adverse event, a change from one health state to another (increasing disease severity), or cost, but instead of using a single methodology to do so, it incorporates several methodologies in fitting a model to a specific data set. For example, the data may be modeled using a combination of regression, correlation, neural nets, decision/classification trees, Markov chains, and other data mining techniques. These methodologies can be used sequentially and in parallel, and the ultimate model that emerges does so from both cooperation and competition among the various methodologies.

Unlike models developed with pure or straightforward methodologies, hybrid models are generally used to fit a specific data set—for example, to predict health care costs in 2005 based on data for health plan members in 2003 and 2004 (as usual, such data may be supplied via claims, demographics, test results, patient input, or the medical record).

The output from a hybrid predictive model is a rule set customized to a specific population for a specific time period. To avoid overfitting the model to the data, the hybrid model is tested against a holdback (validation) data set taken from the same data set from which the development set was drawn (as is conventionally done in all forms of modeling).


Neural networks are artificial simulations of the intricate logical networks of a biological nervous system. It is a simple model composed of a number of independent but interconnected elements (frequently called neurons or units or nodes). Because of their design, neural networks are able to detect subtle interactions between the input variables that affect the output value. For this reason, neural networks have been found to be valuable for pattern recognition, data classification, modeling, and forecasting.

The neurons are very simple models of a biological neuron. Their purpose is to accumulate signals from the neurons in the previous level and decide whether to pass a signal on to the neurons below it. Each neuron makes this determination by applying a unique set of weights to the inputs and calculating the sum. Finally, the neuron transforms this sum to better detect nonlinear patterns and compares the transformed sum with a threshold term to set the output value.

The key feature of a neural network is how these simple neurons are arranged to detect the large and small effects that the input variables have on the output variable. A neural network is typically composed of several layers of neurons with a maze of connections linking the neurons on each layer to those on the layer below it. The most common design in use, a two-layer network, features three layers of neurons: an input layer that accepts the input variables, an output layer that returns the results, and a hidden layer in between.

Once a neural network has been designed, it must be trained on a data set before it can be used for predictions. The training data set includes both the input variables and the associated output variable. Training is an iterative process where each observation is run through the neural network and the prediction is compared with the actual result. Then the weights for each connection are adjusted to reflect the change in experience. Through this process of repeated trial and error, the training process can eventually develop a neural network that is capable of accurately recognizing patterns in data and forecasting future outcomes.

Neural networks are uniquely flexible in their ability to model data. Because of their structure, neural networks will reflect nonlinear relationships between the input variables and the outcome that is being modeled, even when the user is unaware in advance of this relationship’s specific nature. The neurons in the hidden layer allow the network to incorporate a large number of input variables into a nonlinear regression model. This flexibility also leads to concerns about neural networks causing substantial overfitting. Other common concerns with neural networks include difficulty in interpreting the network once it is defined and the speed with which they develop a model.


A prospective predictive model uses information available from one time period to predict an individual’s health care expenditures or to predict the likelihood of certain events in a future period. Prospective models are typically used for predictive purposes—establishing capitation budgets based on predictive claims, for instance. They are also used for the identification of enrollees with potential for case management. In the population health management industry, it is often seen that prospective predictive models are used for predicting who is likely to be high cost next year, who is likely to have inpatient admits next year, or who is likely to develop specific medical conditions such as vascular disease. A prospective predictive model also can be used for return on investment evaluation when a concurrent control group is available.


Clinical rules-based predictive modeling applies rules-based clinical logic to available data from medical, laboratory, pharmacy, and health risk assessment information to find the statistical relationships between current use patterns and future outcomes (i.e., clinical and economic). The predictors could be condition-specific co-morbidities, medication usage, or flags related to compliance issues.


The threshold method is also referred to as a retrospective risk identification method, providing a snapshot in the present time or recent past of a patient’s health status or risk.


Identify members who reach a specific threshold of adverse medical events such as hospital admissions or emergency room visits or a certain amount of health care costs. Threshold modeling methods are based on current health status information and identify current high-risk patients.


Time-series forecasting is a forecasting method that uses a set of historical values to predict an outcome. These historic values, often referred to as a time series, are spaced equally over time and can represent anything from monthly sales data to daily electricity consumption to hourly call volumes.

Time-series forecasting assumes that a time-series is a combination of a pattern and some random error. The goal is to separate the pattern from the error by understanding the pattern’s Trend, its long-term increase or decrease, and its seasonality (the change caused by seasonal factors such as fluctuations in use and demand).

There are two main goals of time-series analysis: (a) identifying the nature of the phenomenon represented by the sequence of observations, and (b) forecasting (predicting future values of the time series variable). Both of these goals require that the pattern of observed time-series data is identified and more or less formally described. Once the pattern is established, we can interpret and integrate it with other data (i.e., use it in our theory of the investigated phenomenon, e.g., seasonal commodity prices). Regardless of the depth of our understanding and the validity of our interpretation (theory) of the phenomenon, we can extrapolate the identified pattern to predict future events.


Duncan, I., and Robb, A. “Population Risk Management: Reducing Costs and Managing Risks”. Intelligent and Other Computational Techniques in Insurance. Hackensack, NJ: World Scientific (2003).

“Neural Network.” 21 Feb. 2006. Wikipedia Free Encyclopedia. 23 Feb. 2006

Patterson, D. “Artificial Neural Networks: Theory and Applications.” Singapore: Prentice Hall (1998).

Warner, B. and Misra, M. “Understanding Neural Networks as Statistical Tools.” The American Statistician Vol. 50 (4) (November 1996).

Predictive Value

Predictive value is a measure of whether or not an individual has a condition based on results of a test, diagnostic process, or algorithm. Positive predictive value is the probability that an individual with a positive test will actually have the disease. Conversely, negative predictive value is the probability that an individual with a negative test truly does not have the condition.

Positive predictive value = True positives / (True positive + false positives)

Negative predictive value = True negatives / (True negatives + false negatives)


Hennekens, C., and Buring, J. Epidemiology in Medicine. Boston: Little, Brown & Company (1987).

Preferred provider organization (PPO)

A preferred provider organization or PPO is a health care delivery system where providers contract with the PPO at various reimbursement levels in return for patient steerage into their practices and/or timely payment. PPOs differ from other health care delivery systems in the way they are financed. They are designed to encourage use of participating network providers, offering better benefits and lower costs for in-network services.

There are two types of PPOs:

  • A non-risk PPO contracts with providers in a geographical area to form an interconnected network of providers and services. The PPO leases its network for a fee to insurance companies, self-insured employers, union trusts, third-party administrators (TPAs), business coalitions and associations.
  • A risk PPO includes a benefit plan and network services either provided by the risk PPO or leased from a non-risk PPO. The risk PPO assumes the financial risk for enrollees’ medical costs.


American Association of Preferred Provider Organizations.


Presenteeism is a term used to describe a situation in which an employee is physically at work but impaired by health problems and not performing to optimal productivity while at work.

See also employee productivity, terms associated with.


Prevalence is a measure of frequency that represents the number or proportion of individuals in a given population who have a particular disease, condition or attribute (e.g., risk factor) over a specified period or point in time. Prevalence refers to all individuals who have a condition, regardless of when the condition started. Prevalence includes both previously identified cases and newly identified cases.

Prevalence contrasts with incidence, which measures newly identified or newly diagnosed cases only. Prevalence is a function of both the incidence of the disease and survival.

Prevalence rate (PR) = Number of cases / Population at risk at the point in time


Patient Time Period
  1/90 1/91 1/92 1/93 1/94 1/95 1/96 1/97  
Patient W
2 yr
Patient X
5.5 yr
Patient Y
4 yr
Patient Z
5 yr
Total years at risk   16.5 yr

Onset of disease

Chronic prevalence: At 1/1/1990, chronic prevalence is one case in two patients, or 50 per 100 (only patients W and Y are eligible at that time). Chronic prevalence at 1/1/1993 is two cases in three patients, or 66.7 per 100. (Patients X, Y, and Z are each eligible at 1/1/1993, and Y and Z have the condition.)


Hennekens, C., and Buring, J. Epidemiology in Medicine. Boston: Little, Brown & Company (1987).

Mausner, J., and Kramer, S. Epidemiology, An Introductory Text. 2nd ed. Philadelphia: Elsevier (1985).

Morton, R., Hebel, J. and McCarter, R. A Study Guide to Epidemiology and Biostatistics. 4th ed. Baltimore: Aspen Publications (1990).

“Overview of Cancer Prevalence Statistics.” Statistical Research and Applications Branch. National Cancer Institute. 23 Feb. 2006

Rothman, K. and Greenland Modern Epidemiology. 2nd ed. Baltimore, MD: Litppinott, Williams & Wilkins (1988).

Primary Payer

This term refers to situations in which a third party payer (e.g., an insurance policy, plan, or program) is responsible for paying first on a claim for medical care. The payer could be private health insurance, Medicare, Medicaid, or another health insurer. What defines them as the primary payer is their role as the first insurer of possibly several to pay the claim, up to their benefit limits. The other payers would then be liable only for any remaining portion of the expenses covered by their benefits. Different rules determine which among private health plans/insurers are considered primary.

See also secondary payer and coordination of benefits.


Glossary: Centers for Medicare and Medicaid. Department of Health and Human Services. 23 Feb. 2006

Program Graduation

The completion of a program (e.g. chronic care management, lifestyle management, case management) or defined series of tasks by a program participant. Graduation is usually accomplished by successfully achieving predetermined time-related or outcomes-based goals. The program may then either disenroll the graduated participant, or reassign him/her to a maintenance program (e.g., one of lesser intensity or a different program), or connect him/her to community resources or their primary care provider. Care should be taken when a graduation is accomplished by predetermined time-related goals, when the patient is disenrolled, without providing follow-up plans; as this may take away the benefits of the program, and endanger life and/or affect quality of life.

In some conditions (for example, asthma), graduation may occur based on length of time since original identification where there is no repeat event, like acute exacerbation, emergency visit, or hospitalization.

Propensity Scoring

A rating or scoring technique that utilizes the likelihood of a decision being made or event occurring for a particular defined group. This tendency for one group or classification is sometimes compared with the relative tendency or likelihood for the same decision or event for another group or classification—for example for the purpose of outcomes measurement or savings estimation.


In an observational study, when there are few covariates (risk factors) in a population that require adjustment, direct adjustment can be made (see actuarial adjustment, for example). This type of adjustment is difficult to perform when there are many different covariates. Propensity scores provide a scalar summary of the covariate information when there are many covariates in a population that require adjustment. The propensity score is a probability that the subject (health plan member, for example) receives a certain treatment. A matched control may be constructed by finding members in the nonintervention population with the same scores as members in the intervention population.


Berg, G., Johnson, A. and Fleeger, E. “Adolescent Telephonic Asthma Care Support Program: A Propensity Score Matched Cohort Study.” Disease Management Health Outcomes. 11 (11) (2003).

Rosenbaum, P., and Rubin, D. “The Central Role of the Propensity Score in Observational Studies for Causal Effects,” Biometrika 70 (1983) 41-55.

Prospective payment

Originally developed for use in treatment of Medicare recipients, prospective payment is a predetermined, fixed amount for a particular service. The amount usually depends on the condition treated or the type of care provided, as defined by a diagnosis-related group (DRG) or a similar categorical system (CPT codes). However, payment rates can be adjusted for outliers--individual cases or services--involving higher utilization, high costs or long stays.


Mosby’s Medical Dictionary, 8th ed. Elsevier, Inc., 2009.

Nash DB, Reifsnyder J, Fabius RJ, Pracilio VP. Glossary of Population Health: Creating a Culture of Wellness. Sudbury, MA: Jones & Bartlett Learning; 2011.

Protected Health Information

Protected health information means individually identifiable health information:

  1. Except as provided in paragraph 2 of this definition, that is:
    1. Transmitted by electronic media;
    2. Maintained in electronic media; or
    3. Transmitted or maintained in any other form or medium.
  2. Protected health information excludes individually identifiable health information in:
    1. Education records covered by the Family Educational Rights and Privacy Act, as amended, 20 U.S.C. 1232g;
    2. Records described at 20 U.S.C. 1232g(a)(4)(B)(iv); and
    3. Employment records held by a covered entity in its role as employer.


67 Fed. Reg. 157 and 45 CFR 160.103.


Protocol is a detailed plan of a scientific or medical experiment, treatment, or procedure.


A protocol is a complex edifice constructed of algorithms. Protocols have the sense of being less flexible than care plans or care paths, though they usually have branching logic to address the likely intermediate outcomes of performing the steps of the protocol.

For example, most chemotherapy is expressed in protocols, which include the types of chemo agents and doses (based on body surface area or weight); what type of and how frequently monitoring for toxicity should be performed; how the dose or timing of chemo should be modified if toxicity appears; what medications should be given for side effects of therapy (e.g., nausea, infection), and under what conditions.

Protocols are generally too rigid for population health management except when care is being directly delivered—for example, in a heart failure clinic. Protocols may also be used to script advice or educational interactions.


The American Heritage Stedman’s Medical Dictionary, New York: Houghton Mifflin Company (2002).