Impact of Therapeutic Switching in Long-Term Care

November 15, 2008
Bryan R. Cote, MA

Elizabeth A. Petersen, MA

The American Journal of Managed Care, November 2008 - Special Issue, Volume 14, Issue 11 SP

Switching medications for nonmedical (formulary) reasons in long-term care settings may increase Medicare Part D resident adverse effects and raise facility downstream costs.

Objectives: To collect clinician perceptions on the impact of switching medications for nonmedical (ie, formulary) reasons in long-term care settings under Medicare Part D and to use the qualitative findings to support more robust research into the effects of Part D policies on nursing home patients and providers.

Methods: We collected individual clinician accounts of experiences with and perceptions about Part D plans when medications were switched for nonmedical reasons, including diseases that clinicians believed were most medically risky for switching with nursing home residents and the common health outcomes of switching. Publisher HCPro’s long-term care readership database was used to reach survey participants. Individuals were classified by clinician type and thene-mailed an invitation to take these electronically disseminated surveys in September 2007.

Results: A total of 569 total case histories representing 555 nursing homes were collected. About three-quarters of the case histories collected showed decreased efficacy after a nonmedical switch for Part D beneficiaries. Switching for those with hypertension, heart disease, diabetes, HIV, and bipolar disorder, among other diseases, was considered most medically risky in this population.

Conclusion: Nurses, physicians, and pharmacists who shared their perceptions regarding the effect of nonmedical switching in long-term care settings reported situations in which switching increased administrative time, increased side effects, and appeared to have increased downstream costs to plans.

(Am J Manag Care. 2008;14(11 Spec No.):SP23-SP28)

Analysis of provider experiences with nonmedical switching policies in long-term care settings indicated that:

  • Residents experienced an increase in adverse effects with their new medications.
  • Pharmacy treatment effectiveness was generally less after a nonmedical switch, putting nursing homes unintentionally at risk for failing to comply with a Medicare quality-of-care regulation.

As growth in the senior population booms toward 40.2 million by 2010, the long-term care (LTC) market is emerging as a more important business line for managed care. Pharmacy and medical spending in the LTC population represents one of the fastest growing categories for Medicare Part D prescription drug plans. For example, of the 150,000 Humana LTC Part D members, approximately 148,000 are enrolled in the company’s stand-alone drug plan (as of August 2007), and costs for this population are 2 to 3 times higher than they are in the retail setting—highest in skilled nursing facilities.1

Given the potential clinical and cost implications of designing and managing a benefit for the LTC population, HCPro, a publisher of regulatory resources for nursing homes, undertook research to understand provider perceptions of and experiences with Part D plan policies. (HCPro publications primarily reach directors of nursing, consultant pharmacists, medical directors, nurse practitioners, and administrators who use the publisher for its regulatory resources and Medicare training.) The goal of this research was threefold:

1. Identify major challenges and Part D policies, and their effects on providers and patients in skilled nursing facilities.2. Collect individual facility accounts through a qualitative survey that could be used as an impetus for more robust research and policy discussions.

Initial Questionnaire and Phone Interviews

To uncover more case histories and facility accounts as to whether outcomes for nursing home patients and day-to-day quality of life could be adversely affected when a Part D plan switches medications for nonmedical reasons, we developed a follow-up electronic survey. Our hope was to gather individual facility profiles and then analyze them to see whether there were any noteworthy trends. At the time of the research in September 2007, little information existed to show whether switching for nonmedical reasons in the LTC population led to any negative outcomes (eg, an increase in hospitalizations).

Three online surveys—one each for nurses, physicians, and pharmacists—were created to identify the prevalence of nonmedical switching and to learn what each provider group’s experience has been. Surveys contained the same questions generally. (See the eAppendix Table at for a list of the questions disseminated to the participants.) In each survey, we defined switching as therapeutic substitution; we further alerted each clinician filling out the survey that switching did not refer to substituting generic versions of the same molecule.

The initial survey respondents (n = 884) were invited to take the follow-up survey during the week of September 5, 2007. To ensure that they were qualified professionals and had the relevant experience beyond the minimum 10 years in LTC, HCPro asked each to fill out a pre-survey screening checklist to help identify their background in LTC. The checklist included these items:

• Minimum of 10 years’ experience in LTC setting.• Membership in the major LTC associations (American Medical Directors Association, Association Directors of Nursing Administration/Long Term Care, American Society of Consultant Pharmacists).

• High understanding of F-Tag guidelines for nursing homes (based on rating scale).

• Experience with 3 Part D plans.

• At least 2 years working at or consulting for current facility (required).

The researchers only reviewed responses from individuals meeting 6 of the 8 criteria. Other information also was collected prior to the qualitative questions (ie, respondent’s title, location of facility).

The HCPro researchers acknowledged several limitations in collecting perceptions and individual case histories. For example, it was difficult to ask clinicians to recall their experiences from 2005 (before the Part D drug benefit took effect). To address this recall limitation, the research questionnaires asked respondents to provide general insights on the prevalence and implications of switching in this population in 2007 compared with before implementation of Part D.


More than three-quarters of the respondents (76.6%) said it was common for a resident’s new drug to be less effective after a switch for formulary reasons. In a nursing home, this is relevant for Medicare compliance reasons. For example, a portion of the clinicians noted circumstances in which a Part D plan’s formulary preference (the therapeutic switched drug) was tried and proven ineffective. As a result, these facilities were forced to dispense a drug ineffective for the patient, and by doing so risked a violation of Medicare’s F-Tag #329 unnecessary drug guidelines (see Discussion section).

Side Effects

Table 2 shows the differences between what nurses, pharmacists, and physicians “perceive and recall” as the most common negative outcomes after a nonmedical switch at the time they took the survey in 2007. Although there are limitations in asking for opinions without backup claims data, these findings suggest that potential quality of care and possibly cost implications are associated with these policies. More than half of the physicians said they are at least occasionally called on to prescribe new medications for symptoms or side effects that arise as a result of nonmedical switches, though only about one-fifth of pharmacists thought so. Sixty-two percent of physicians said nonmedical switching resulted in increased side effects, compared with 49% of pharmacists and 24% of nurses. It’s difficult to understand why the majority of nurses surveyed did not believe that side effects increased after a nonmedical switch, when most said that the new medication was less effective. This disconnect may be due in part to a lack of education on recognizing various signs and symptoms of diseases.

Switchbacks: Plan and Patient Health Impact

Nurses thought that hypertension was the most medically risky disease in which to make a switch for nonmedical reasons. Physicians said depression was most risky, and pharmacists said depression and bipolar disorder were the most risky. Highlights from the clinician case experiences include:

• 88% of LTC residents with a gastrointestinal/stomach disorder were somewhat likely to highly likely to experience reduced efficacy if their medication was switched for nonmedical reasons.

• 100% of residents with Parkinson’s disease were somewhat likely to highly likely to experience reduced efficacy if their medication was switched for nonmedical reasons.

Staff Time

Although the individual facility accounts suggest that nursing homes could be experiencing some increases in medical, pharmacy, and administrative costs—and ultimately costing Part D plans—how widespread and important these increases are could not be extrapolated from these anecdotal findings. However, according to one respondent, the findings serve as a cautionary tale: “Some Medicare Part D plans don’t understand anticholinergic side effects in the elderly, and their formularies illustrate their lack of education on this issue” (Malcolm Fraser, MD, President, Bay Geriatrics, phone interview of September 18, 2007). The possibility of adverse events and other negative outcomes increases as patients age, so what might be appropriate for a 65-year-old might cause an unwanted consequence in an 85-year-old.

Part D: Before and After

The survey and interview responses suggest that more exploratory research is needed on the effects of therapeutic switching in nursing home patients receiving the Medicare Part D benefit. Given the number of comorbid conditions for many patients and the higher risk of drug–drug interactions and safety accidents related to side effects, the responses from the 569 survey participants serve as cautionary guidance for plans. The Part D payer community or individual Part D plans may need to address their policies further as their LTC membership increases. According to comments from several of the physician survey respondents, switches within classes are not a major issue. However, if a Part D plan requires a resident to switch medications out of a chemical class, doing so brings inherent risks and potential downstream costs.1 For example, anticonvulsants have a narrow therapeutic index, and switching usually will require drug-level monitoring.1

On another level, medication-switching policies in this population have created an educational disconnect. Here is an obvious example cited by Lesha Yerka, RN, Director of Nursing, Eden Park Nursing Home, in a phone interview on September 21, 2007:

We have doctor’s secretary/nurse refuse to assist with nursing home prior authorizations that are needed when switching is required by the plan. They think the nursing home can do it. We need to communicate with the MD, but are not given a chance to do so. We also have a problem with the MD who receives 2 prior authorizations, since a resident needs 2 pills with different doses to make up the ordered dose. The MD signs one, but not the other so we have to repeat the process. Also, the state is not accepting the date the drug was ordered and filled. They are paying from the day they receive the PA signed by the MD. So the nursing home gets stuck with the bills for meds that are ordered because we have to give them, but the MD feels too overworked to do the paperwork for the orders. We had this happen recently over a sleep medication; the MD got the PA and he changed the name of the drug and the PA was denied. We got stuck with paying for 50 pills to the tune of $224.00. This whole process takes up a lot of my time in a week, but doesn’t help to decrease cost especially if you include my salary time.

At the same time as LTC providers have been managing Part D, nursing homes have had to comply with stricter Medicare medication monitoring guidelines that took effect in January 2007. Respondents complained that Part D switching policies conflicted with a Medicare regulatory mandate called F-Tag #329, which requires LTC facilities to ensure that any prescription medication is helping a patient to reduce side effects and meet the goals of their therapy—not produce the opposite effect. For example, if a drug causes side effects (eg, pain that prevents a patient from taking part in occupational therapy), nursing homes have a regulatory mandate under this F-Tag #329 to respond by adjusting the drug’s dose or making a change in therapy or intervention.

Complying with the F-Tag #329 mandate was becoming difficult for some facilities in the face of Part D therapeutic substitution policies. The fact that some nursing directors report conflicts between Part D switching policies and the FTag requirements is cause for discussion among stakeholders and perhaps greater education across the LTC continuum—payers, providers, and patients. The timing of F-Tag #329 and Part D presents an opportunity for managed care to conduct true medication therapy management for LTC.1


2. Centers for Medicare & Medicaid Services. nursing home compare web site. Accessed September 17, 2008.