Our rates of adherence for depression screening were remarkably low as reported in the provider's encounter note and the indicator demonstrated good inter-rater reliability. In the Pacific Northwest VHA, the majority of patients are screened for depression by a primary care intake nurse but the results are not reported in encounter notes. Our study only extracted provider encounter notes. The provider must be aware of the screen results and feel the results are clinically important to make a notation in their encounter note. Thirty-five percent of the PD patients in our sample had been given an ICD-9 CM code for depression and the prevalence of depression in other PD patient populations has been shown to be 40–50% [8
]. Therefore, it is likely that providers are either 1) not determining the screening results of their patients or 2) not recording the presence or absence of depression in their PD patients' management. Although encounter visits doubled after patients began seeing specialists, the rate of annual screening for depression did not improve. Given that depression is so prevalent in PD patients and that depression and disability are associated [10
], our findings strongly suggest that providers are neither recording the presence of depression nor considering depression an important co-morbidity in PD patients. Interestingly, patients with non-VA care had nearly double the rate of depression screening recorded in the encounter note compared to those without outside care. Since we did not review outside notes, this increased rate reflects VA providers noting the presence or absence of depression in their own evaluation and management. The higher rate of depression screening in patients with outside VA care could be for several reasons. Patients with outside care may be more pro-active and demand their providers be more thorough in their management. VA providers may be also more thorough because they worry that patients with outside care may not be getting as good as care as patients without outside care.
Adherence to recommended fall screening was also very low and the indicator demonstrated good inter-rater reliability. The most common complications of patients with PD are axial features such as gait impairment and falls [11
]. In a prospective study of PD patients with average duration of illness of three years, falls occurred in over two-thirds of PD patients annually with half of the PD patients experiencing more than one fall [12
]. For 100 PD patients followed over two months, 20 hospitalizations occurred for falls of which three-quarters were for fractures and the rest for lacerations [13
]. Finally, falls are a leading cause of nursing home admissions in a general elderly population [14
]. Given the prevalence of falls and its association with poor patient outcomes, the low adherence rates for screening are concerning.
Patients who received care outside the VA had higher screening rates for falls and this may be for the same reasons as the higher depression screening rates seen in patients receiving outside care. However unlike depression screening, patients with specialty care had higher rates of adherence for annual fall screening. There are two possible reasons why patients who are seeing specialists have a higher rate of adherence to annual fall screening. It may be that patients that see specialists are more likely to have fallen and therefore report falls to their providers. Alternatively, it may be that specialists are more aware of falls as a complication of PD and are more likely to ask patients about falls. Regardless, considering that even with specialty care, adherence rates were less than a third, interventions specifically targeting fall screening should be of prime importance.
Discussion of treatment options for urinary incontinence had the highest adherence rates of any of the indicators and the indicator also demonstrated good reliability. Greater than half of patients who had symptoms of incontinence received appropriate management advice. Surprisingly, patients who received specialty care were much less likely to receive appropriate care. The high adherence rate among non-specialty providers may reflect the fact that urinary incontinence occurs with some frequency in patients without PD. Nevertheless, urinary symptoms occur in 40% of PD patients [11
]. Our finding that patients who received specialty care in addition to primary care had lower rates of adherence suggests that there may be a negative outcome from the coordination of care between specialists and generalists. Both provider groups may believe that the other provider group is responsible for managing urinary incontinence.
Inter-rater reliability for the management of hallucinations could not be calculated due to not enough instances of an assessment recorded as positive to reliably report management. Of those patients who had hallucinations, greater than half were managed appropriately. The presence of hallucinations has been shown to be a stronger predictor of nursing home placement for patients with PD compared to motor and cognitive impairment [15
]. In particular, the presence of hallucinations early in the treatment of parkinsonism is associated with increased probability of nursing home placement [17
]. The low number of eligibility periods we found in this study suggests that providers may not be asking or recording the presence of hallucinations. Because of the importance of management of hallucinations for patient outcomes, future research may want to first focus on accurate screening for hallucinations.
Inter-rater reliability for the management of orthostatic hypotension also could not be calculated due to not enough instances of an assessment recorded as positive to reliably report management. Adherence was less than a third and none of the predictor variables influenced adherence.
In addition to evaluating rates of adherence to several evidence-based care indicators in patients with PD, this study had two other important findings. Our study confirmed that medical chart review is necessary to identify cases of PD. Less than half of the patient medical charts identified with administrative data had confirmed or probable PD on medical chart review. The author has shown previously that administrative data is inaccurate in identifying cases of PD for quality of care assessments [7
]. This study also found that only half of the veterans with PD were seen by specialists. This is consistent with other studies of non-VA PD populations in both the U.S. and Canada that have shown PD patients have limited access to specialty care [3
Our study is limited in that two of the care indicators had too few instances of an assessment recorded as positive to reliably report management and therefore caution must be taken in interpreting the results of those indicators. For the rest of our indicators and for qualifying diagnoses, inter-rater reliability was high. While blinded, both our raters were involved in the development of the abstraction instrument and therefore their familiarity with the instrument may have resulted in higher rates of reliability than might otherwise be found in raters.
Another limitation of our study is that our rates of adherence to PD care indicators may not be generalizable outside the Pacific Northwest VA system. Rates reported by the VA Greater Los Angeles Healthcare system using a similar instrument were higher than those reported by us: annual screening rates for depression of 67% and annual screening rates for falls of 52% [19
]. The differences in rates between the two VHA systems may be a result of the higher prevalence of PD specialty care at the VA Greater Los Angeles Healthcare system or the fact that care was evaluated in the year 2003 at West Los Angeles VA whereas care was evaluated from 1998–2004 for the Pacific Northwest VA. Regardless, these findings suggest a large amount of geographic variability of care delivered to PD patients and require further evaluation.
The findings of this study suggest that physician professional societies or Parkinson's disease organizations might consider promoting screening for PD patients to improve adherence to care indicators. These organizations could make available a standardized set of questions about recent falls, hallucinations, symptoms of orthostatic hypotension and a self-administered depression scale that PD patients could complete prior to their clinic visits.