Initial pharmacotherapy in a population of veterans with Parkinson disease.

Since the advent of the first dopamine agonist (DA), controversy has existed regarding the choice of medication for the initial therapy of PD. While levodopa (LD) remains the most potent medication to relieve the cardinal symptoms of PD, the disabling complications, dyskinesias and motor fluctuations, may be prevented by initial therapy with a non-LD medication.

Recent clinical trials demonstrating the delay in onset of motor complications in patients have led to treatment initially with DAs, especially in young-onset PD patients. The effectiveness of monoamine oxidase B (MAO-B) inhibitors, amantadine, catechol-O-methyl transferase (COMT) inhibitors, and anticholinergic drugs in treating the cardinal features of PD has given health care providers further options when selecting initial therapy. However, all these medications may impair cognitive function, and there is insufficient evidence that initial treatment with MAO-B inhibitors, amantadine, or COMT inhibitors prevents or delays the onset of motor complications. Since primary care physicians diagnose and manage a large percentage of patients with PD, we hypothesized that the specialty designation of the prescribing health care provider may influence initial therapy.

Discussion

Neurologists are initiating less than one-third of the antiparkinsonism therapy for patients with PD. Our finding is consistent with other studies showing that neurologists see the minority of patients with PD. LD as initial treatment outweighs other medications and is mainly prescribed by primary care providers. Younger patients were more likely to receive a DA than other antiparkinsonism medications, but many younger patients with PD are not being started on a DA.

Our regression model shows that after considering the effect of patient characteristics, an MDS would be more likely to prescribe a DA than a non-MDS neurologist, suggesting important differences in care practices among neurologists based on subspecialty expertise.

After considering the effect of provider specialty and patient characteristics, DAs were also more likely to be prescribed in the nonconsecutive years 2000 and 2003, suggesting an organizational or professional influence, such as the distribution of organization-wide memos or the publication of clinical trials.

Mental health providers overwhelmingly prescribe anticholinergics for initial therapy of PD. This may be because they are more familiar with anticholinergic medications and therefore default to this category of medications for the initial therapy of PD. This finding is concerning because patients on anticholinergics were more likely to have a diagnosis of dementia, and anticholinergics are known to worsen cognition.

The strengths of using VHA administrative data to understand provider prescribing behavior is that the patient population is selected independent of the provider or clinic and therefore not biased by referral selection. The limitation is that identifying patients using ICD-9 codes and pharmacy data may not identify all cases; however, the cases are likely to be representative of the population. Further, these findings may not be generalizable outside the VHA.

After considering patient differences, initial pharmacotherapy for PD is strongly influenced by the provider specialty designation. Because initial therapy is influenced by nonclinical factors, many clinically eligible patients with PD may not be receiving therapy that could delay the onset of motor complications.