American Geriatrics Society (AGS) 2013 Annual Scientific Meeting

May 3-5, 2013; Grapevine, TX
___________________________________________________________________________________________________

Age Does Not Independently Predict Poor Outcomes in Patients With Escherichia coli Infection

It is well known that older adults are more susceptible to healthcare-associated infections than younger adults, and that bacterial infections have become a therapeutic challenge for long-term care clinicians given the rise in multidrug resistance. Extended-spectrum beta-lactamases (ESBLs) are enzymes that produce resistance to most beta-lactamase antibiotics, including penicillin. CTX-M is a subtype of ESBLs, which has been associated with high cost and mortality in hospitals, long-term care facilities, and acute care units. A study presented during a poster session at the AGS annual meeting focused specifically on the epidemiology of CTX-M–producing Escherichia coli in older adults, finding that older age alone is not independently associated with poor clinical outcomes among persons with ESBL-producing E coli infections. The study was conducted by Sameen Farooq, student, Wayne State University, MI, and colleagues, Detroit Medical Center (DMC), MI.

The team used a retrospective matched case-controlled method to gather data about the clinical isolates of ESBL-producing E coli collected from all patients at DMC over a 16-month period. Patients were divided into two age groups: those aged 65 years and older (n=232), and those between the ages of 18 and 65 years (n=145). According to the patients’ medical records, all patients were given effective therapy, which was based on in vitro antimicrobial susceptibility results as reported by the DMC microbiology laboratory. Next, they performed bivariate analyses to create a propensity score for predicting the likelihood of being an older adult as opposed to a younger adult; and after controlling for the propensity score, multivariate regression analysis was performed to determine the independent impact of older age on clinical outcomes. The bivariate analysis showed that older age was significantly associated with high mortality and longer length of hospitalization. However, neither of these outcomes was associated with age in the multivariate analysis.

“The multivariate analysis accounts for the confounding variables and shows that age itself is not related to higher mortality and longer length of hospitalization,” said Farooq in an interview with Clinical Geriatrics^®. “The analysis further highlights the notion that older patients are at higher risk for infections and worse outcomes because of comorbidities, decreased immunity, and higher exposure because of more traveling between hospitals and community settings—not simply because of older age.”

The researchers observed that ESBL-producing E coli were isolated frequently from cultures taken from younger adults in the emergency department, indicating that there is a significant presence of CTX-M type ESBL-producing E coli in the community. Farooq said that this signals a pressing healthcare concern because “at this point, the systems of care have not developed effective protocols to augment treatments and lessen the burden of disease for at-risk populations. This aspect of the study warrants further analysis as it speaks to the nature of modulating infection-control protocols in the face of an ever-changing landscape. For now, there is a need to better understand the epidemiology of E coli and focus on development of treatment and management protocols for hospitals, long-term care facilities, and the broader community setting.”—Allison Musante
__________________________________________________________________________________________________________________

Medication Reconciliation Intervention Yields Low Impact on Reducing Rate of All-Cause Readmissions of Elderly COPD Patients

Hospital readmissions have been an enduring problem for healthcare providers, but they have become a national concern recently in light of the Hospital Readmissions Reduction Program. Under the terms of the program, which went into effect in October 2012 as part of the Affordable Care Act, hospitals incur a financial penalty for readmissions of patients with heart failure, acute myocardial infarction, and pneumonia to the same hospital within 30 days of discharge. In 2015, readmissions for patients with chronic obstructive pulmonary disease (COPD) will also be reported to the Centers for Medicare & Medicaid Services; thus, many administrators are looking ahead at ways to reduce readmissions in this large patient population.

According to Christine Eisenhower, PharmD, University of Rhode Island College of Pharmacy, and colleagues, more than one-third of patients aged 65 years and older who present to the emergency department are hospitalized because of an adverse drug event (ADE). To reduce the risk of all-cause readmission in this group, Eisenhower and colleagues speculated that ensuring medication reconciliation by a pharmacist at hospital discharge can decrease medication discrepancies resulting in ADEs. The team presented the results of their study during a poster session at the AGS annual meeting.

The study included 60 men and women over the age of 65 years with a history of COPD who were admitted for any cause to the New Hanover Regional Medical Center and Cape Fear Hospital between January 31, 2012, and February 29, 2012. At hospital discharge, a pharmacist conducted a face-to-face medication-clarification meeting with the patient regarding his or her home regimen and orchestrated medication reconciliation using the patient’s charts and electronic orders; any communication discrepancies between patients and providers were also resolved. At the 30-day follow-up, six patients had died and 13 patients were readmitted for any cause.

Although 12 medication discrepancies were identified and reconciled, Eisenhower and colleagues found that the pharmacy services did not impact readmission rates. According to baseline data from March 2011 for patients aged 65 years and older with COPD, the rate of 30-day all-cause readmission at both facilities was 21.1%; the average cost per hospitalization was approximately $10,722; and the average length of stay was 6.2 days. At the end of the 30-day study period, the readmission rate had increased by 0.6% from the previous year. They noted that the average length of stay per patient decreased from an average of 7.25 days for the first hospitalization to 7.17 days for the second hospitalization. Additionally, the cost per hospitalization decreased by approximately $1455, but it was not clear if the change could be attributed to the intervention.

The authors acknowledged that the study was limited by a short intervention period, unknown medication compliance, unknown outpatient follow-up after discharge, and small sample size. They concluded that future studies can better assess the impact of a full-time pharmacist to reduce all-cause readmissions if the study “adds patient medication counseling at discharge, tracks outpatient and medication compliance, [and] works with prescribers to select affordable medications.” Furthermore, they noted, similar studies can expand the intervention process to include other chronic disease states.—Allison Musante
______________________________________________________________________________________________________

Patients Taking Medications on the 2012 Beers Criteria Are at Risk of Longer Hospitalization

Since its inception in 1991, the Beers Criteria for Potentially Inappropriate Medication Use in Older Adults has been an important evidence-based resource for healthcare providers working to reduce the risk of adverse drug reactions (ADRs) and ADR-associated hospitalizations. Dozens of studies have assessed the clinical value of the Beers Criteria, which have been revised in 1997, 2003, and most recently in 2012, to optimize patient medication safety in a variety of care settings. A new study has compared the utility of the 2003 Beers Criteria to the 2012 Beers Criteria in reducing risk of ADRs and reducing length of stay (LOS) in acute care. The study was presented at the AGS annual meeting by Victoria Huang, student, Department of Pharmacy Practice, Rutgers, the State University of New Jersey, and Timothy Reilly, PharmD, Department of Clinical Pharmacy, University Medical Center of Princeton (UMCP), NJ.

Huang and Reilly conducted a retrospective chart analysis, which included men and women aged 65 years and older who were admitted to the UMCP acute care unit between January 20, 2011, and May 20, 2011. Patients were divided into two groups: a control group, in which none of the patients were using a drug from the 2012 Beers Criteria list (n=111), and an experimental group, in which all of the patients were using one or more drugs from the 2012 Beers Criteria list (n=229). The same patient population was evaluated in a 2011 study by Callinan and Reilly, except that the 2003 Beers Criteria was used.

Huang and Reilly found that the average LOS in the control group was 5.32 days, compared with 7.99 days in the experimental group (P=.00082), a finding that was consistent with the previous study. However, unlike the other study, Huang and Reilly reported no statistically significant evidence that use of a drug on the Beers Criteria affected the risk of an ADR (P=.182). In an interview with Clinical Geriatrics^®, Reilly said these data should be carefully interpreted because the lists of drugs included in the 2012 Criteria and the 2003 Criteria are different. For example, Reilly and Huang found that the most commonly prescribed drug from the 2012 Beers Criteria (16%) was an antipsychotic, whereas Callinan and Reilly found ferrous sulfate (>325 mg/d) to be the most commonly prescribed drug on the 2003 Beers Criteria. Ferrous sulfate was removed from the 2012 Beers Criteria.

Reilly added that he and his colleague are in the process of reanalyzing the data by taking into account potential confounders, such as disease severity, to get a truer sense of the rate of ADRs in users of Beers Criteria medications.

“Considering that a significant number of the patients we evaluated were admitted by geriatricians, we were somewhat surprised to see the prevalence of Beers Criteria medication use in our study population,” said Reilly. “The rate that we observed was higher than previously published data, although those datasets used an older version of the Criteria and were consistent with our prior analysis of this dataset using the older version. This highlights that we can certainly improve on our medication use practices in the elderly, but also that there are some situations where a provider must use a Beers Criteria drug, particularly in patients with multimorbidity.”

Reilly emphasized that both studies provide evidence that patients taking medications on the Beers Criteria list have a significantly longer LOS, putting older adults at increased risk of hospital-associated complications. Moving forward, he said, “The Beers Criteria is an important document to guide prescribing in the elderly and a good place to start when looking at medication use in this population, but it certainly isn’t the only or best resource to use; there is no substitute for sound clinical judgment.”—Allison Musante
________________________________________________________________________________________________________

Prescribers Advised to Move Away From the Sliding Scale

Sliding scale insulin (SSI) regimens approximate daily insulin requirements by increasing the preprandial or nighttime insulin dose using predefined blood glucose ranges, a practice that has been reported to have numerous disadvantages and to place patients at high risk of adverse drug events (ADEs). As a result, the 2012 AGS Updated Beers Criteria for Potentially Inappropriate Medication Use in Older Adults included SSI as a potentially inappropriate medication for elderly persons with type 2 diabetes; however, a poster presented at the AGS annual meeting found that healthcare professionals working in skilled nursing facilities (SNF) still frequently order SSI regimens and that these regimens are likely to cause hypoglycemic events.

The poster, which was presented by researchers affiliated with North Shore-Long Island Jewish Health System, NY, outlined the results of a 1-year retrospective randomized chart review of 100 SNF patients (mean age, 81 years; range, 65-95 years) with type 2 diabetes. Most patients were female (56%) and white (88%). The average weight was 162 lb (range, 76-256 lb), with a mean body mass index of 27 kg/m² (range, 15.9-52.7 kg/m²). Patients had moderately impaired function, an average comorbidity index of 5, and hospital and SNF stays of 6 and 18.5 days, respectively. The chi-squared test or Fisher’s exact test were used to compare the demographics and characteristics between patients receiving (n=57) and not receiving (n=43) SSI regimens. In addition, Poisson regression with adjustment for overdispersion was used to model glycemic events in both patient populations.

A total of 38 episodes of hypoglycemia (blood glucose, <60 mg/dL) were reported, with most events occurring between 4 am and 8 am. The hypoglycemic episodes involved 18 patients, 13 of who were on an SSI regimen; eight were on aggressive regimens and five were on conservative regimens. The researchers found that many of the patients on the SSI regimen were also receiving sulfonylureas (28%), basal insulin (17.5%), and basal and preprandial insulin (17.5%).

The study revealed that patients receiving SSI regimens were 2.6 times more likely to have a hypoglycemic event than their non-SSI counterparts, with a hypoglycemic event incidence of 2.28 per 100 SNF patient days versus 0.86 per 100 SNF patient days, respectively. The investigators, which included Judith Beizer, PharmD, a member of the expert panel of the AGS 2012 Beers Criteria, concluded that the study supports the AGS approach to diabetes management and that increasing awareness of the AGS 2012 Beers Criteria is essential to improve the quality of medical care for older patients with diabetes.

Clinical Geriatrics^® (CG) had the opportunity to discuss the study and SSI in greater detail with three of the study researchers: Santiago Lopez, MD, the principal investigator of the study and a geriatric fellow; Liron Sinvani, MD, hospitalist/geriatrician; and Gisele Wolf-Klein, MD, director of geriatric education.

CG: Can you please define conservative versus aggressive SSI use?

Dr. Lopez: SSI regimens were divided into aggressive and conservative categories based on a study by Queale and colleagues, which used glucose levels of <175 mg/dL and >175 mg/dL, respectively.

Dr. Sinvani: To further elaborate, the terms conservative versus aggressive are meant to represent the way SSI regimens are ordered; meaning at what starting glucose level the insulin therapy is initiated and the amount of insulin given initially and for each incremental increase in glucose level. For our project we chose to base our definition on Queale’s study, as Dr. Lopez mentioned.

You heavily assessed demographics and patient characteristics. Did you identify any factors that increased the likelihood of a patient being placed on an SSI regimen?

Dr. Wolf-Klein: We did not identify any factors that increased the likelihood of a patient being placed on an SSI regimen. We also found that none of the diseases in the patients’ medical histories, with the exception of liver disease (P=.049), were significantly associated with hypoglycemic status in those who received SSIs.

Why do you think SSI regimens are still so commonly used, despite widely reported drawbacks and the recent update to the AGS Beers Criteria?

Dr. Lopez: Historically, the SSI system has been taught and used by multiple generations of physicians over many decades. It has been the “standard of care” for diabetic patients in hospitals, in nursing homes, and in their own homes. What we have seen in the literature is that the prevalent use of SSI regimens is attributed to convenience, simplicity, and its ability to provide prompt treatment. The sliding scale ensures that insulin therapy will be given when hyperglycemia is first recognized. The regimen is easily implemented in general surgical and clinical areas, and it does not depend on locating the attending physician or other staff concerning the necessary insulin dosage.

Dr. Sinvani: SSI regimens are predominantly used in acute and subacute care. SSI use is especially rampant in the hospital setting. Patients with type 2 diabetes take numerous types of oral hypoglycemics. On admission to the hospital, oral hypoglycemic regimens are switched to SSI regimens for a variety of reasons, including that the SSI drugs are on the hospital’s preferred drug list; patients need to undergo tests requiring fasting or going off the unit; and convenience to medical staff, who would otherwise need to do finger sticks around the clock to determine treatment. Although awareness of the dangers of SSI regimens is increasing, there still seems to be a general lack of knowledge on this front and of the new Beers Criteria. In addition, for those aware of the drawbacks, there is no structured alternative that is widely known or used. Therefore, SSI regimens are used by default. Educating healthcare professions and further research are needed for widespread change to occur.

What insulin protocol do you recommend that physicians employ when treating elderly patients who have type 2 diabetes?

Dr. Wolf-Klein: The pioneer work of Van den Berghe published in 2006 demonstrated for the first time that tight control of glycemia in the ICU setting was detrimental to diabetic patients, contrarily to the general practice and the beliefs of physicians. We are now learning to better individualize the insulin treatment of our older diabetic patients, basing treatment on their comorbidities, age, glycated hemoglobin levels, life expectancies, functional status, and residential setting.

Dr. Sinvani: In treating elderly patients with type 2 diabetes, it is extremely important to treat each patient on an individual basis. Insulin use in the elderly can be difficult because of their decreasing functional status (ie, ability to inject insulin), comorbid conditions, and polypharmacy. In addition, there are natural age-related changes in body composition and in kidney, pancreatic, and liver function to contend with. The most important thing, as in all of medicine, is to do no harm. Hypoglycemic episodes, especially in the elderly, can be detrimental. Therefore, if insulin is needed to treat an elderly patient, a conservative approach that takes into account the patient’s functional status, blood glucose control, comorbid conditions, eating patterns, overall prognosis, and preference is essential to proper care.—Christina T. Loguidice