E D I T O R' S B L O G | S E P T E M B E R 1 9, 2 0 1 7 | 3 : 3 2 P M Comparing aspects of care delivered by the integrated neurologist
Kevin A. Kerber, MD, MS, Ann Arbor MI
A challenge for the primary care medical home model (PCMH) is incorporating care from specialties. In this study, Elrashidi et al explored the impact of integrating a neurologist into the Mayo Clinic PCMH by comparing aspects of care delivered by the integrated neurologist with matched visits from the pre-integration time period. The integrated neurologist worked 0.6 FTE, had a staff of 3 FTE registered nurses, 3 clinical assistants, and a specified work plan -- 3-4 scheduled patients per half day, in addition to curbsides, e-consults, follow-up, and additional acute same-day consults. Over a 12-month period of patient follow-up, patients seen by the integrated neurologist had fewer subsequent neurologist visits (0.62, p=0.001), EMGs (OR 0.64, p=0.009), and brain MRIs (0.60, p<0.001), but not total subsequent outpatient visits (0.92, p=0.21), ED visits (0.83, p=0.20), or hospitalizations (0.96, p=0.83) compared with the pre-integration visits. The integrated neurologist also did not lower the time to neurologist appointment (p=0.83), despite presumably reducing referrals because of curbsides and e-consults.
This is a provocative idea, but unfortunately the study has too many limitations to render a judgment on the effect of integrating neurologists into PCMHs. There are simply too many potential confounders relating to the study’s design. The major design limit is the evaluation of only one integrated neurologist. Potential confounders include differences in salaries, staff support, volumes of patients, time with patients, and unmeasured patient factors (e.g., severity, care preferences). It would have been more informative to quantify how this one neurologist’s practice changed after the integration, although that addresses only one limitation. Realizing however that it is not feasible for most organizational changes to undergo large randomized trials, is there some data from this work that could better inform the potential effects of an integrated neurologist? This is a circumstance where semi-structured interviews with key persons (the neurologist, office staff, primary care providers, and patients) and qualitative analysis might provide important insights. For example, I would be interested in the key stakeholders description of their practice experience before and after integration.
Lesli E. Skolarus, MD, MS, Associate Professor, Department of Neurology, University of Michigan Health System
Health equity, as defined by The Robert Wood Johnson Foundation, means that “everyone has a fair and just opportunity to be as healthy as possible. This requires removing obstacles to health such as poverty, discrimination, and their consequences, including powerlessness and lack of access to good jobs with fair pay, quality education and housing, safe environments, and health care.” The first step to obtaining equity in neurologic care is identifying health disparities, work that was recently undertaken by Altaf et al. This group of researchers analyzed the national state of equity in neurologic care using data from the 2006–2013 Medical Expenditure Panel Survey (MEPS), a nationally representative household survey, which queries medical conditions as well as medical expenditures, including outpatient and inpatient visits. After adjustment for socio-demographics, insurance status, health status, and self-reported neurologic diagnosis, the researchers found that black participants were nearly 30% less likely to see an outpatient neurologist and Hispanic participants were 40% less likely to see an outpatient neurologist compared to white participants.
As the authors discussed, there are a number of limitations to this work, including the reliance on self-reported surveys, which may introduce response and recall bias. Additionally, the assumption that all racial and ethnic groups have the same neurologic need is unlikely to hold true. For example, blacks have higher incidence of stroke, and greater post-stroke disability than whites, suggesting that certain populations have a need for greater healthcare utilization. Finally, research has not determined when or whether encounters with a neurologist result in better, neutral or worse health outcomes
Despite these limitations, Altaf et al. took the important first step of beginning the exploration of racial disparities in neurologic care. I look forward to their continued work as they dive deeper into this topic. As the nation is contemplating healthcare reform, it is important to keep in mind that health equity requires additional resources to the populations who face the greatest social and health challenges, not the reduced efforts as currently proposed.
E D I T O R' S B L O G | J U N E 1 6, 2 0 1 7 | 3 : 2 9 P M Diabetic peripheral neuropathy pain and pharmacotherapy
Brian C. Callaghan, MD, MS, University of Michigan Health System
Waldfogel et al performed a systematic review of pharmacotherapy for diabetic peripheral neuropathy pain. Similar to four previous systematic reviews, they conclude that serotonin reuptake inhibitors (SNRIs), tricyclic antidepressants (TCAs), and pregabalin are effective medications for this highly prevalent condition. In contrast to previous studies, they conclude that gabapentin is not effective while oxcarbazepine is effective.
For gabapentin, this is an important distinction since it’s in the same class as pregabalin (calcium channel ligands), but much less expensive (~$30 vs. $300 per month). While the standardized mean difference (SMD) for gabapentin was not significantly different than placebo, the magnitude was larger than pregabalins. Furthermore, the class I studies from the AAN guideline showed that the effect size of gabapentin and pregabalin were similar. Moreover, Finnerup et al’s systematic review revealed that the number needed to treat (NNT) for gabapentin is lower than that for pregabalin (7.2 vs. 7.7). Taken together, the difference in evidence for these two drugs is negligible, and also illustrates that a p value <0.05 is an arbitrary cutoff. Similar to TCAs and SNRIs, drugs in the calcium channel ligand class likely have similar effects on neuropathic pain (class effect). For oxcarbazepine, the SMD is significantly different from placebo, but two of the three trials were negative trials and all three were deemed class II studies by the AAN. Further research is needed to clarify whether oxcarbazepine is effective. The authors further state that botulinum toxin is effective therapy for diabetic neuropathic pain. However, the evidence is very weak. Only two trials have been performed, both with fundamental flaws (small and no placebo response). Better trials are needed to determine whether botulin toxin is effective. Finally, the authors found that typical opioids are not effective, but atypical opioids (tramadol, tapentadol) are effective treatments. Given the mounting evidence of increasing morbidity and mortality with opioid treatment, this provides even stronger reason to avoid typical opioids in the treatment of chronic neuropathic pain. Atypical opioids are different from typical opioids in that they have SNRI like effects. This may explain the difference in efficacy, but the question remains whether the benefit outweighs the long term side effects. Overall, this systematic review supports the conclusions of previous studies that SNRIs, TCAs, and calcium channel ligands have the strongest evidence in neuropathic pain whereas opioids should likely be avoided.
E D I T O R' S B L O G | A P R I L 5, 2 0 1 7 | 1 1 : 5 5 A M Hawkins and Berman provide an excellent summary of the pharmacology of the novel anti-psychotic pimavanserin
Roger L. Albin, MD, University of Michigan
Psychosis is a common problem in advancing Parkinson disease (PD). There are significant limitations of standard approaches of adjusting dopamine replacement therapies and using unconventional anti-psychotic agents such as clozapine and quetiapine. Pimavanserin is a promising addition to our therapeutic arsenal. Hawkins and Berman thoughtfully emphasize our relatively limited knowledge of pimavanserin clinical pharmacology. It was approved on the basis on 1 trial in a selected group of PD participants. Additional trials, particularly in PD dementia and Lewy body dementia subjects, and good post-marketing surveillance are needed to fully understand the utility of this agent.
How should clinicians use this agent now? Psychotic symptoms, usually visual hallucinations and illusions, are common in PD but frequently don’t require treatment. Like other PD features, psychotic symptoms should be treated when they cause disability. Many PD patients experience infrequent, non-threatening hallucinations. Some patients have preserved insight into the unreality of their hallucinations, mitigating the need for treatment. In some demented patients, even frequent hallucinations do not cause significant disability. Pimavanserin, like other new agents directed at niche markets, is expensive. As Hawkins and Berman point out, costs to patients will vary considerably. Regardless of costs to patients, the social cost remains relatively high. After some discussion, our Movement Disorders group decided to treat pimavanserin as a second line agent. We elected to initially trial low dose quetiapine in PD patients with troublesome psychotic symptoms. If unsuccessful, we try pimavanserin. Our approach may change as experience with pimavanserin accumulates.
Practicing clinicians, especially Movement Disorders specialists, can play an important role in expanding our knowledge of pimavanserin effects. The discovery of dopamine agonist related impulse control disorders, for example, began with a small case series reported by Mark Stacey. Dr. Stacey’s example is a good model for clinicians using new agents.
For more information access the full article, Pimavanserin: A novel therapeutic option for Parkinson disease psychosis.
E D I T O R' S B L O G | J A N U A R Y 2 5, 2 0 1 7 | 4 : 2 5 P M Dr. Lyell K. Jones Answers Our Questions on the Axon Registry
Kevin A. Kerber, MD, MS, Ann Arbor MI
In Introducing the Axon Registry, Sigsbee et al outline the rationale, structure, function, and challenges related to the AAN's development of its own clinical quality data registry: the Axon Registry. Dr. Lyell K. Jones, co-author, answers some questions we had after reading the report.
Q1) The article mentions that registries guide improvements at the population level. What are the best examples of quality metric registries leading to meaningful improvements?
While we have a great deal of experience with registries in medicine, clinical quality data registries are relatively new. Several other specialty societies have led the way in the developing registries similar to the Axon Registry, and have learned a great deal about the quality of care delivered to their respective patients. One example is the American College of Cardiology’s PINNACLE registry, which has yielded numerous insights into measure performance, clinical outcomes, and disparities in cardiovascular care. Many readers of Neurology may be familiar with the American Heart Association’s Get With the Guidelines registry, which has provided cerebrovascular neurologists with extremely valuable data regarding stroke outcomes and real-world effectiveness of cerebrovascular interventions. Below the system and population level, individual practices that participate in registries like Axon have an unprecedented window in their processes and outcomes of care, and can use their own data to improve quality of care.
Q2) Is it a problem that the entities creating the quality registries are also the ones the most vested in demonstrating value?
The foremost goal of the Axon Registry is to provide a tool for neurologists to examine and then improve the quality of care they provide to patients with neurologic disease. A prerequisite for this goal is that the data housed in the registry are accurate and valid. The AAN has a robust data governance policy that safeguards the integrity of the data and ensures that access to the data for analysis is done responsibly, with full disclosure of potential conflicts. There certainly will be opportunities at the system and database level for independent health services researchers to use Axon data to examine elements of the value equation, and we encourage them to do so.
Q3) The article mentions that validation steps are performed to assure the accuracy of data capture. Can you share with us the reliability thresholds for concluding measure performance as accurate?
The Axon Registry validation strategy includes practice level validation, systematic reviews of measure logic, and validation at the database level. One of the primary 2017 goals for the Axon Registry will be completing our database level validation. Given the scope and complexity of that work, our plan is to work with an external vendor to complete the validation process. Many specific details such as reliability thresholds either haven’t been determined or are too preliminary to share broadly.
For more information, see Introducing the Axon Registry: An opportunity to improve quality of neurologic care,