How applicable are clinical practice guidelines to elderly patients with comorbidities?

Definitions of main outcome measures

The main outcome measures (ie, mention of older adults or people with comorbidities, time needed to treat in the context of life expectancy, and barriers to implementation) were selected from previous studies that examined applicability of CPGs to elderly patients.^13,14

We defined elderly as people older than 65 years of age. When no specific age was indicated, we looked for words that signified older age in the CPGs (eg, senior, older adults, frail elderly).

The applicability of CPGs to people with comorbidities was determined by scanning CPG documents for any mention of people with 1 or more chronic conditions in addition to the primary disease that would affect treatment choices.

The outcome of time needed to treat in the context of life expectancy has also been used in previous literature.^13,14 It refers to specific recommendations on how to modify treatment for the elderly in the context of life expectancy. The recommendations must balance the risks and benefits to the patient; specifically, the guidelines should address the treatment within the context of life expectancy, prognosis, quality of life, or end-of-life care.

Finally, the outcome measure of barriers to implementation includes explicit statements that help clinicians to understand the limitations that might pose difficulty in executing the recommendations.

Data collection and analysis

Relevant data were extracted using a data collection sheet. The data collection sheet was constructed from existing validated instruments, including the Appraisal of Guidelines for Research and Evaluation instrument^20,21 and a checklist for reporting CPGs developed by the Conference on Guideline Standardization,²² and other published studies that focused on applicability of CPGs to elderly patients with multiple comorbidities.^13,14 The final data collection sheet included 4 key indicators: any mention of older adults, mention of people with comorbidities, time needed to treat to benefit in the context of life expectancy, and barriers to implementation of the CPG. Data were collected, summarized, and tabulated in an Excel spreadsheet.

Recommendations specific to the elderly

Table 4^{24–28,31–38} shows examples of statements and recommendations about issues related to the elderly in CPGs. With the exception of the dementia^32–38 and congestive heart failure²⁸ (CHF) CPGs, the rest of the CPGs specified the applicable age in their recommendations. Two CPGs (diabetes²⁶ and CHF²⁸) used words such as elderly or frail elderly. The diabetes²⁶ and depression³¹ CPGs provided evidence that treatment response was similar in older people and in the younger population. Three CPGs (diabetes,²⁶ hypertension,²⁷ and depression³¹) provided a list of suitable and unsuitable pharmacotherapy and nonpharmacologic therapies to be used in elderly patients.

Recommendations for people with comorbidities

Table 5^{23,25–28,30–38} provides a summary of CPGs that mentioned or addressed people with comorbidities. Two CPGs (CHF²⁸ and depression³¹) recommend multidisciplinary and interdisciplinary care of patients with multiple comorbidities. In addition, the diabetes²⁶ CPG recommends a multidisciplinary and interdisciplinary approach to all diabetes patients, regardless of age. The depression³¹ CPG provides a list of medications with low potential for drug-drug interactions. On the other hand, the osteoarthritis²³ CPG leaves it to a physician’s discretion for how to handle comorbidities. The chronic obstructive pulmonary disease³⁰ (COPD) CPG provides different treatment regimens for those with and those without comorbidities. Finally, the diabetes²⁶ CPG recommends modification of glycemic targets among elderly patients with comorbidities. In addition, some CPGs provide recommendations on modification of treatment targets (diabetes²⁶ CPG), comorbidites that warrant screening for lipid profiles (dyslipidemia²⁵ CPG) or comorbidity-specific treatment recommendations (hypertension²⁷ CPG).

Time to benefit in the context of life expectancy

Table 6^26,28,30,31 shows how several CPGs addressed issues related to treatment in the context of life expectancy. Four CPGs (CHF,²⁸ COPD,³⁰ diabetes,²⁶ and depression³¹) discuss issues related to end-of-life care. The CHF,²⁸ COPD,³⁰ and diabetes²⁶ CPGs recommend modification of treatment goals in accordance with prognosis and patient choice. The depression³¹ CPG recommends indefinite treatment of elderly patients with partial resolution of symptoms.

Barriers to implementation

Table 7^{26–28,31–38} provides a summary of CPGs that list or discuss barriers to implementation. Most of the listed CPGs acknowledge gaps and a lack of good evidence for recommendations pertaining to the elderly population with or without comorbidities. Barriers to implementation are discussed in terms of adherence, polypharmacy, and difficulty in achieving treatment targets.

Evidence levels of recommendations

Recommendations in each guideline were generated by using different evidence grading systems. We list the available evidence level of the recommendations targeting elderly with comorbidities in Tables 3^{23–28,30–38} to 7.^{26–28,31–38} The recommendations varied from grade A to grade D.

Elderly with or without comorbidities

We reviewed CPGs for 10 common chronic conditions managed by primary care physicians in elderly patients. This study has shown that most of these CPGs do mention the elderly population (7 out of the 10 CPGs, which is comparable to US¹³ and Australian¹⁴ studies). However, while most of the CPGs mention elderly patients, few of them adequately discuss issues related to elderly patients with comorbidities. Only 3 CPGs (CHF,²⁸ diabetes,²⁶ and depression³¹) have dedicated sections addressing elderly patients with comorbidities. The osteoarthritis²³ and osteoporosis²⁴ CPGs provide recommendations specific to elderly patients, but the recommendations are broad and nonspecific. The osteoarthritis²³ and dementia^32–38 CPGs have minimal discussions on issues pertaining to elderly patients with comorbidities. The dementia^32–38 CPGs recommend that physicians stop medication if there is high risk of drug-to-drug interaction, although they did not specify common comorbidities that providers should be aware of.

Time needed to treat in the context of life expectancy

Only the CPGs for CHF,²⁸ diabetes,²⁶ and COPD³⁰ specifically recommend modification of treatment in the context of life expectancy and patients’ preferences and prognoses. The other CPGs do not address issues related to time needed to treat to benefit in the context of life expectancy. This finding might be partly explained by the fact that primary studies often focus on evaluation of benefits rather than harms, which requires much more expensive longitudinal studies.³⁹ Recently, one study focused on developing a payoff time framework, which might identify patients for whom particular clinical guidelines are unlikely to confer benefit; however, this framework has not been validated in a clinical setting.⁴⁰

Barriers to implementation

Only half of the reviewed CPGs provide explicit statements on when or under what circumstances CPGs might be difficult to implement. Such statements address issues like treatment targets in the elderly, warnings about polypharmacy or drug-drug interactions, and the lack of evidence for some recommendations. Acknowledgment of barriers to implementation is an important component of continuous quality improvement of CPGs. By identifying and explicitly stating areas where it might be difficult to implement CPGs, strategies can then be developed to address the shortcomings and therefore improve the quality of CPGs or develop strategies to overcome the barriers.

Evidence level of recommendations

We found that very few recommendations for elderly with comorbidities were based on clinical trials (ie, grade A recommendations). This is not surprising, as most randomized trials exclude the elderly, especially those of advanced age or with comorbidities.³⁹ It is estimated that only 5% of trials are designed specifically for elderly patients 65 years of age and older, and 15% of trials exclude elderly participants without any justification.⁴¹

Limitations

This is the first Canadian study that has attempted to study applicability of CPGs to elderly patients with comorbidities. Our study has several limitations. This study evaluated only 10 CPGs. The conditions discussed are the most common chronic conditions seen in the elderly and are managed largely by primary care physicians. Other diseases such as cancer and thyroid problems were not included but would likely be in the realm of both specialist and primary care providers. Second, there is no validated instrument to evaluate applicability of CPGs to elderly patients with comorbidities. We constructed our data collection sheet from different sources. We covered only 4 items that we believed were specific to the elderly and those with comorbidities. It is likely that there are other important domains that were not evaluated in this study. Future studies should develop and validate an instrument for assessing applicability of CPGs to elderly patients with comorbidities.

Conclusion

This study shows that only a handful of CPGs adequately address important issues common in the care of elderly patients. Given the demographic transition occurring in Canada, this is of urgent importance. Clearly, there is considerable room for improvement in these CPGs. There is a pressing need to improve the evidence base that undergirds the care of people of advanced age and with multiple concurrent chronic diseases.⁴² Guideline developers should include much more detailed information on management of elderly patients, with a particular emphasis on eliciting patient and caregiver concerns, setting clinical priorities, managing expectations (particularly around prognosis), and fostering optimum communication. This will aid in making CPGs more patient centred rather than disease driven. Based on this research, we propose that an ideal guideline should consider an open discussion about patients’ preferences, benefits of intervention in advanced age, time to benefit from treatment, trade-offs for function over disease control, as well as acknowledgment of uncertainty.