An Assessment of Three Post-Discharge Drug-Based Quality Assurance Events for Elderly Medicare Beneficiaries Enrolled in a Stand-Alone Part D Plan

An important aspect of continuity of care is hospital discharge planning. Hospital discharge planning is needed for all patients, but especially for elderly patients due to their increased prevalence of chronic conditions and drug consumption.(1,2) One of the key components of discharge planning involves the performance of medication reconciliation by hospital personnel.(3,4) Medication reconciliation refers to the process of resolving discrepancies in a patient’s medication record to identify the most comprehensive and accurate list of current medications taken by the patient at the time of discharge. During medication reconciliation, prescriptions received prior to the hospitalization should be reconciled with all new prescriptions issued at discharge.(5) Medication reconciliation helps guard against inadvertent medication errors such as therapeutic duplication and drug-drug interactions, which can lead to unplanned readmission.

For some patients, a medication reconciliation follow-up beyond the immediate discharge may be needed to reinforce/reaffirm previously reconciled medications.(6) This follow-up can be critical for elderly patients who suffer from cognitive impairment as well as lower visual and/or hearing acuity.(7) The introduction of new medication regimens, especially for patients experiencing cognitive impairment, coupled with the potential lack of caretakers, adds to the vulnerability of elderly patients for a medication mishap. Additionally, spousal caretakers may be overwhelmed by their new responsibilities post-discharge; with many competing accommodations to consider (e.g., follow-up appointments, basic needs, and arrangement of home health), medication management concerns may receive less attention. Thus a medication reconciliation follow-up in the community setting should be considered to ensure that there is proper continuity of medications and assurance that any other medications not mentioned at the time of the hospitalization are accounted for.

The present study assesses the occurrence of three critical quality assurance events (QAEs) in the community setting for 30 days following hospital discharge for elderly Medicare beneficiaries enrolled in stand-alone Part D plans. The three QAEs are drug–drug interaction, drug–disease interaction, and therapeutic duplication. Our primary study aim is to examine the distribution of drug-based QAEs post discharge across five-day increments. Our second aim is to identify characteristics associated with post-discharge QAEs. The study provides useful information to practitioners and policymakers in identifying the most effective timing of medication reconciliation post-discharge in the community setting.

Methods

Data

Enrollment and claims data for 2009 through 2011 were obtained from the CMS’ Chronic Conditions Data Warehouse (CCW). The CCW is a research database that includes data for Medicare fee-for-service beneficiaries and Part D prescription drug event data for Medicare Part D enrollees. To assess the three QAEs, the First Databank classification system (proprietary to First Databank; license held by CMS) was applied to the CCW data to identify drug–disease interactions, drug–drug interactions, and therapeutic duplication.

Study Sample

The study sample consisted of elderly beneficiaries (≥ 65 years) with medication use, living in the United States, and enrolled in a stand-alone Part D program for all 12 months of 2010. We excluded beneficiaries residing in a nursing facility, those with a skilled nursing facility stay, those who died within 30 days of a hospital discharge, and those having two hospitalizations within a 30-day period. Our study sample of 1,625,292 beneficiaries represented about 6% of the Medicare Part D population of approximately 28 million enrollees.(8)

Measures

A QAE was considered present if the beneficiary experienced at least one drug–drug interaction, drug–disease interaction, or therapeutic duplication during the 30-day period post-discharge. QAEs included: (1) events that existed prior to the admission and reoccurred post-discharge; and (2) events that were newly present following the discharge. Drug–drug interactions exist when one medication alters the metabolism of another medication (e.g., metronidazole increases the anticoagulation effect of warfarin). Drug–disease interactions exist when an underlying disease alters the metabolism of a drug. An example would be giving a COX-2 selective nonsteroidal anti-inflammatory drug to a beneficiary with chronic kidney disease, potentially leading to decreased renal clearance. Therapeutic duplications exist when two drugs (e.g., beta-blockers labetalol and metoprolol) of the same medication class are given simultaneously. The concern is double dosing, because additive effects or drug potentiation (i.e., the combined effect is greater than the additive effect) can occur. For drug–drug interactions and therapeutic duplications, we required that: (1) the base drug be refilled before its day supply was depleted; and (2) the intervening drug be dispensed during the original fill timeframe of the base drug. This requirement was used to increase the likelihood that the base drug was being taken at the same time that the intervening drug was given (i.e., an intervening drug/base drug combination event occurred). We captured prescription fills up to 90 days before 2010 and up to 90 days following 2010 to account for 90-day prescription fills.

Study sample characteristics included the use of home healthcare services following discharge, sex, age, race/ethnic origin, census region, disease burden, Medicare and Medicaid dual eligibility status, metropolitan area, and prescription burden. Sample characteristics were based on the beneficiary’s enrollment status in 2010. Age, disease burden, and prescription burden were coded as categorical variables. Age was divided into the groups of 65 to 74, 75 to 84, and 85 years of age and over. Disease burden and prescription burden were categorized as low, medium low, medium high, and high. Disease burden was assessed based on the 2010 CMS Hierarchical Condition Category risk score data. Prescription burden was assessed based on the number of unique medications (i.e., by generic name) utilized by the enrollee during 2010.

Analysis

We performed a descriptive, cross-sectional analysis at the enrollee level, reporting each of the three QAE measures separately in five-day increments for the 30-day period post-discharge. An additional analysis using a logistic regression model was performed to assess enrollee characteristics associated with a QAE (reported in aggregate, and not separately) occurring during the 30 days. Regression results were reported as odds ratios (OR) along with p values. Analyses were performed using SAS 9.4 software (proprietary to SAS Institute, Inc., license agreement held by CMS).

Results

The prevalence of QAEs among enrollees in our study sample was highest with drug–disease interactions (11%), followed by therapeutic duplications (5%) and drug–drug interactions (< 1%) (Table 1). Of the 30 days observed post-admission, the initial five-day window following discharge contained the highest proportion of QAEs. The number of enrollees affected was almost double for the first five-day window as compared with the second, third, fourth, and fifth windows. These findings suggest that the initial window (and days 2 and 3 in particular) following discharge is a “sweet spot” when medication problems are most likely to occur. However, there is also a notable increase in QAEs near the 30-day limit following discharge. This increase is seen particularly in drug–drug interactions and with therapeutic duplications.

Regression results (Table 2) indicate that the use of home health services following discharge reduced the likelihood of a QAE occurring (OR = 0.91, p <.001). Women were 28% more likely to have a QAE following a hospital discharge compared with men (OR = 1.28, p <.001). The oldest elderly (≥85 years) were 30% less likely to have a QAE following a hospital discharge compared with the youngest elderly (age 65–74) (OR= 0.7, p <.001). Blacks and Hispanics were 23% and 12% less likely, respectively, to have a QAE following a hospital discharge compared with whites (OR = 0.77, p <.001 and OR = 0.88, p <0.001, respectively). Enrollees with a high disease burden were 2.16 times as likely as those with a low disease burden to have a QAE following a hospital discharge (OR = 2.16, p <.001). Dual enrollees were 27% more likely to have a QAE following a hospital discharge compared with non-duals (OR = 1.27, p <.001). Urban status enrollees were 4% more likely to have a QAE following a hospital discharge compared with their rural counterparts (OR = 1.04, p <.001). Lastly, enrollees with a high prescription burden were 2.62 times as likely as those with a low prescription burden to have a QAE following a hospital discharge (OR = 2.62, p <.001, respectively).

Discussion

This study looked at the distribution of QAEs during the first 30 days post-discharge for elderly Medicare beneficiaries. The results indicate that the best time to identify and fix medication problems is as soon after discharge as possible, even though patients presumably undergo medication reconciliation just prior to hospital discharge. Beneficiaries at greatest risk for a post-discharge QAE are female, 65 to 74 years in age, white, have dual enrollment in Medicare and Medicaid, have urban residence, and have higher disease and prescription burdens. Beneficiaries at lower risk of a post-discharge QAE received home healthcare following discharge.

A recent study found that more than half of adverse drug events occurred within the first 14 days following discharge.

The relative prevalence of the QAEs suggests that computerized drug systems currently in place are performing correctly by catching drug–drug interactions and therapeutic duplications at the time of prescribing. These computerized systems use medication names and fill dates captured at the pharmacy counter to identify potential drug–drug interactions and therapeutic duplications quickly. However, these same computer systems are hampered in their ability to identify drug–disease interactions due to insufficient information regarding beneficiary disease states.

Our findings are consistent with prior research. A recent study found that more than half of adverse drug events occurred within the first 14 days following discharge.(9) Our findings differ in that an even more compressed timeframe for addressing medication issues was found to exist—specifically, within two to three days of discharge.

However, we also noted a 30-day surge in QAEs following discharge. This finding suggests that the incorporation of new medications or the changing of existing medications is a common occurrence that can ultimately lead to confusion for the beneficiary regarding which medication(s) to take. This surge may result from an influx of new medications via automatic refills of medications prescribed around the time of discharge. New medications also may be prescribed during post-discharge office appointments.(10) Office visits often result in new medications, considering that the writing of a prescription is the medical intervention most frequently used by a prescriber.(11,12) For example, consider the prescribing of a new medication to treat an adverse reaction associated with a previously prescribed medication—an event known as the prescribing cascade.(12,13) What had been a balanced medication regimen at the time of discharge slowly starts to become unstable as the beneficiary sees both current and new physicians.

Medication reconciliation is considered a key process in the provision of home health services post-discharge.

We found that the likelihood of QAEs decreased when home health services were involved. Medication reconciliation is considered a key process in the provision of home health services post-discharge.(14) Beneficiaries who are in poor health and have low incomes have relatively higher rates of home health use.(15) Medication issues should be more apparent in this population due to their frailty, and increased vigilance by home health assistance may prove helpful.

Our findings suggest that all beneficiaries would benefit from receiving professional services both immediately following hospital discharge as well as a month following discharge. This finding aligns with the care transition models advocated by Coleman and Naylor,(16) in which a transitional care nurse visits the beneficiary’s home within a day or two of discharge, followed by periodic visits providing guidance to the beneficiary. CMS introduced two new procedural billing codes that may encourage Medicare providers to perform medication reconciliation following a hospital discharge. Effective in 2013, the transitional care management codes reimburse providers for care management services, including medication reconciliation, following a hospital discharge.(17) Effective in 2015, the chronic condition management code provides non–face-to-face care coordination services 24 hours per day, seven days per week to beneficiaries with two or more chronic conditions.(18) The code requires billing providers to develop a comprehensive care plan that includes medication management, conduct post-discharge follow-up, and perform medication reconciliation with review of adherence and potential interactions. Future research will address uptake of these services and whether they are successful in addressing some of the gaps identified in this paper.

QAEs following discharge continue to be a challenge for the healthcare system. When the responsibility for a beneficiary’s care transfers from the watchful eye of hospital staff to the home setting, vulnerability to medication discrepancies increases. To reduce this vulnerability, a more proactive approach is warranted. Mechanisms must be devised that allow healthcare providers to reach beneficiaries sooner (following discharge). Enhancing pharmacist involvement in the care transition process post-discharge may be another option for consideration.(19-21)

Limitations

Our findings are based on cross-sectional data and therefore are suitable for identifying associations between measures and beneficiary characteristics, but not causal relationships. Additionally, QAEs categorized as problematic may, in fact, be appropriate following a thorough evaluation by the prescribing practitioner. Thus the QAEs found in our study are suggestive but not necessarily definitive. Also, the study sample may not be fully representative of the broader Part D population. For example, in comparison to the general Part D population over 65 years old, the study sample was skewed toward the older elderly (≥75 years of age) as compared with the youngest elderly (65–74 years of age), whereas gender and race were similar (data not shown). Furthermore, the age of our data may be a limiting feature considering that the impact of CMS’ two new procedural codes (addressed previously) are not captured in our data. Lastly, the low R-squared in the regression model indicates that not all factors contributing to a QAE are being captured. R-squared is a statistical measure of how well the regression model fits the data.

Conclusion

To enhance continuity of care in the elderly, performing a medication reconciliation in the community setting as soon as possible following hospital discharge should be encouraged. The reconciliation preferably should occur within two to three days following discharge—a much shorter timeframe than previously identified but consistent with the care transition models.

References

1. Garcia-Caballos M, Ramos-Diaz F, Jimenez-Moleon JJ, Bueno-Cavanillas A. Drug-related problems in older people after discharge and interventions to reduce them. Age Ageing. 2010;39:430-438.

2. Walker C, Hogstel MO, Curry LC. Hospital discharge of older adults. How nurses can ease the transition. Am J Nurs. 2007;107(6):60-70.

3. Meguerditchian AN, Krotneva S, Reidel K, Huang A, Tamblyn R. Medication reconciliation at admission and discharge: a time and motion study. BMC Health Serv Res. 2013;13:485.

4. Paparella S. Medication reconciliation: doing what’s right for safe patient care. J Emerg Nurs. 2006;32:516-520.

5. Greenwald, JL, Halasyamani L, Greene J, et al. making inpatient medication reconciliation patient centered, clinically relevant and implementable: a consensus statement on key principles and necessary first steps. J Hosp Med. 2010;5:477-485.

6. Bellone JM, Barner JC, Lopez DA. Postdischarge interventions by pharmacists and impact on hospital readmission rates. J Am Pharm Assoc. 2012;52:358-362.

7. Gray SL, Mahoney JE, Blough DK. Medication adherence in elderly patients receiving home health services following hospital discharge. Ann Pharmacother. 2001;35:539-545.

8. Medicare Part D population comprised of Prescription Drug Plan enrollees, Medicare Advantage Prescription Drug Plan enrollees, and Retiree Drug Subsidy enrollees. Medicare & Medicaid Statistical Supplement, 2011. www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/MedicareMedicaidStatSupp/2011.html . Accessed March 14, 2016.

9. Kanaan AO, Donovan JL, Duchin NP, et al. Adverse drug events after hospital discharge in older adults: types, severity, and involvement of Beers Criteria Medications. J Am Geriatr Soc. 2013;61:1894-1899.

10. Mankey CG, Varkey P. Discontinued medications: are they really discontinued? Agency for Healthcare Research and Quality, Patient Safety Network. psnet.ahrq.gov/webmm/case/325/discontinued-medications-are-they-really-discontinued . Accessed March 10, 2016.

11. Cherry D, Lucas C, Decker SL. Population aging and the use of office-based physician services. NCHS data brief, no 41. Hyattsville, MD: National Center for Health Statistics. 2010; www.cdc.gov/nchs/data/databriefs/db41.pdf . Accessed May 5, 2016.

12. Rochon PA, Gurwitz JH. Optimising drug treatment for elderly people: the prescribing cascade. BMJ. 1997;315:1096-1099.

13. Rochon PA, Gurwitz JH. Drug therapy. Lancet. 1995;346:32-36.

14. Fuji KT, Abbott A. Ensuring effective medication reconciliation in home healthcare. Home Healthcare Nurse. 2014;32:516-522.

15. Ellenbecker CH, Samia L, Cushmah MJ, Alster K. Patient safety and quality in health care. In: Hughes RG. Patient Safety and Quality: An Evidence-Based Handbook for Nurses. Rockville, MD: Agency for Healthcare Research and Quality. www.ncbi.nlm.nih.gov/books/NBK2631/ . Accessed March 10, 2016.

16. Health Policy Brief: Care Transitions. Health Affairs. www.healthaffairs.org/healthpolicybriefs/brief.php?brief_id=76 . Accessed March 10, 2016.

17. CMS: Medicare Learning Network. Transitional Care Management Services: ICN 908628. www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/Downloads/Transitional-Care-Management-Services-Fact-Sheet-ICN908628.pdf . Accessed August 1, 2016.

18. CMS: Medicare Learning Network. Chronic Care Management Services: ICN 909188. www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/Downloads/ChronicCareManagement.pdf . Accessed August 1, 2016.

19. Kristeller J. Transition of care: pharmacist help needed. Hosp Pharm. 2014;49:215-216.

20. Kripalani S, Roumie CL, Dalal AK, et al. Effect of a pharmacist intervention on clinically important medication errors after hospital discharge. Ann Intern Med. 2012;157:1-10.

21. Walker PC, Bernstein SJ, Tucker Jones JN, et al. Impact of a pharmacist-facilitated hospital discharge program. Arch Intern Med. 2009;169:2003-2010.