Twenty-seven studies were included in the review (n=1,405). Six RCTs were included (n=481, sample size range 10 to 255). Three RCTs assessed levodopa infusion (n=50) and three assessed deep brain stimulation (n=431). Randomisation and allocation concealment were reported to be adequately described on three RCTs. The other studies were uncontrolled studies. Follow-up across all studies ranged from six days to four years.
Outcomes related to motor complications:
In the three RCTs that assessed levodopa infusion, the percentages of a patient's waking day in a functional on state ranged between 80% and 100% at the end of each study. In two RCTs that evaluated deep brain stimulation (surgery), positive functional outcomes were reported in terms of percentage increases in on times (66% and 138%). Statistically significant differences were reported for both RCTs in favour of surgery/immediate surgery compared to best medical therapy (p<0.001). The direction of findings was confirmed in uncontrolled studies. Three non-controlled apomorphine studies found a decrease in off time from baseline to the end of the studies.
The UPDRS motor score from one RCT of levodopa infusion showed an improvement from 25.5 at baseline to 14.5 at the end of the study (absolute change was not stated). Change in activities of daily living score was from 16 to 14. Results from uncontrolled studies of levodopa infusion were inconclusive and those that assessed apomorphine infusion showed no effect. Two of the three RCTs of deep brain stimulation showed improvements in UPDRS motor scores (28% and 41%) and activities of daily living scores (69% and 39%); the third and largest RCT reported decreases in UPDRS (29%) and activities of daily living (24%) scores. Results from uncontrolled studies were inconclusive.
Patient-related quality of life:
One RCT of levodopa infusion showed a statistically significant (p<0.01) difference in median end of study Parkinson's Disease Questionnaire (PDQ) summary index score (25) compared to conventional treatment (35). One non-controlled levodopa study showed a statistically significant (p<0.005) improvement in the PDQ summary index score from 59.5 at baseline to 49.2 at end of study. Improvements of between 17% and 24% in PDQ summary score were reported for three RCTs of deep brain stimulation, two of which reported significantly better improvements for deep brain stimulation compared to conventional treatment (p=0.02 and p<0.01).
Adverse events and withdrawal rates were sparsely reported. Adverse events included administration site problems, dyskinesia, somnolence, hallucinations and nausea.