Population health: What’s medical necessity got to do with it?

March 8th, 2017 / By Barbara Aubry, RN

A busy doctor just saw his diabetic patient in the office. The usual was done; B/P, weight, pulse, respirations recorded and the blood glucose log was reviewed. The doctor reminded his patient to see podiatry for his diabetic foot check and ophthalmology for his eye exam. The doctor did a general exam, then gave his patient a new Rx for his diabetes medications and told him to make a follow-up appointment for next month. Routine for the provider – not so routine for the patient. What was not asked may result in an ED visit for diabetic crisis. Why? Because the patient’s wife lost her job; they cannot afford to keep their car because they cannot afford the insurance on his disability benefits alone. So, he has no way to fill or pick up his prescriptions. Since he was not asked, he is too embarrassed to volunteer his family troubles even though his wife urged him to “talk to the doctor.”

Medical necessity criteria

You all know this but I feel obliged to list it again as a review. Medical necessity is the key concept used by payers to describe any healthcare intervention that is:

  • Recommended by a healthcare professional
  • A treatment for a clinical condition
  • Provided in an appropriate level of care
  • Efficacious in improving health outcomes as determined by scientific evidence, accepted care standards of medical practice or expert opinion
  • Cost-effective for the patient’s condition when compared to alternative interventions
  • Not solely for the covered person’s convenience or the convenience of the patient’s family or provider

Documentation of our patient’s encounter with his healthcare provider met the principals of medical necessity as is generally understood. Until recently this was considered adequate documentation but with the adoption of alternate payment models that are largely acuity-based, knowledge of the patient’s total risk is essential to provide appropriate intervention and protection of provider revenue.

Social determinants and third party reimbursement

Full disclosure – I used to be a case manager. If you asked most ED docs and hospital-based case managers or social workers if social determinants impacted patient outcomes the answer would be a resounding “yes!” But even though we were aware of the unmet social needs, at that time we believed little could be done other than trying to find either charity services or state or federal programs that could offer some support.

Since one of the main pillars of medical necessity states “coverage is not available for convenience of the patient, patient’s family or provider” the industry did not expect third party payers to consider reimbursement for “convenience” even if it resulted in higher patient care costs and was a major contributor to poor outcomes. Some bold not-for-profit hospitals cobbled together transportation services for their low-income patients who could not attend clinical appointments without access to transportation support. But these programs were few and far between and had little to no reliable funding.

As data is more available and work focuses on reducing costs, third-party payers are beginning to recognize trends that can be managed. For example, data shows low-income individuals are more frequently admitted to the hospital or visit the ED at months end for either diabetic or hypoglycemic crisis. This happens as they run out of money for food and medicine.  Efforts can focus on providing emergency access to food and medicine to reduce high-cost hospital use. Third-party payers focusing on these types of interventions can reduce their costs while improving outcomes. But more data is needed to link social issues with health outcomes. Is it possible to support and advance the tenants of population health while working within the current documentation and coding framework?

Recording social determinants

Many of you are familiar with the Evaluation and Management Documentation Guidelines, both the 1995 and 1997 versions. Currently, there is a section under the “history” area called the “past family, and/or social history” (AKA PFSH) area. It is generally understood that this area is reserved for medical or surgical family history. The “social” history includes an age-appropriate review of past and current activities. These likely include marital status, smoking and drinking behavior, employment, but historically have not been used to document non-medical determinants.

In my opinion, providers need to be encouraged to add documentation regarding the following in order to support the move to third-party reimbursement of non-medical symptoms that have direct detrimental impact on health costs and outcomes of entire populations:

  1. Problems related to education and literacy – how can a patient follow directions if he/she can’t read?
  2. Problems related to employment and unemployment – if our provider had asked his diabetic patient this question, perhaps he would have been able to connect the patient with available resources to help
  3. Occupational exposure to risk factors
  4. Problems related to housing and economic circumstances
  5. Problems related to social environment
  6. Problems related to upbringing
  7. Family circumstances
  8. Problems related to certain psychosocial circumstances

The list above represent options in ICD 10-CM under the heading “Persons with potential health hazards related to socioeconomic and psychosocial circumstances”, code range Z55-Z65.

Documentation changes in support of improved population health

This is not intuitively difficult, but it does require expanding the litany of questions asked of patients – and not necessarily by clinicians. CMS trained “navigators” to assist individuals who were signing up for coverage under the ACA since the requirements could be confusing. Why not use a similar navigator functionality to collect data from patients that is not clinically-based? This data gather can be performed in the hospital, clinic and/or provider’s office. For example, document the patient’s reply to questions such as:

  • Do you have access to transportation for your doctor appointments?
  • Are you able to fill your medication prescriptions?
  • What is your living arrangement? How many stairs?
  • What was the reason you were unable to fill a recent prescription?
  • Do you have reliable access to food every day of the month?
  • Who is dependent upon you for care?
  • Who are you dependent upon for care?
  • Do you have a family member with a drug, alcohol or psychiatric condition?

Answers to the questions above become part of the medical record and should be reviewed by the clinician – and that review should be officially documented in the medical record. Any changes in treatment should reference the problems it impacts. Using the appropriate ICD-10 code for social determinants should be as important to the provider and payer as the medical diagnosis. This supports data mining that can directly impact actions to create new programs to manage the social issues that result in higher healthcare utilization with poor outcomes. It also helps both providers and payers understand the real risk and acuity of the population they serve.

Documenting all patient problems encountered supports both frequency and intensity of services. In addition, it supports Medicare’s determination of medical necessity largely through the thorough documentation of the patient’s needs. This is likely to result in better outcomes on audits, plus it provides payers with additional data required to support creative programs to meet non-medical needs that impact medical outcomes.

Barbara Aubry is a regulatory analyst for 3M Health Information Systems.