Assignment 4

In patients with Parkinson’s disease, does Lee Silverman Voice Training (LSVT) BIG Program demonstrate more of a decrease in motor symptoms than traditional

physical therapy?

Assignment #4

December 14, 2015

In patients with Parkinson’s disease, does Lee Silverman Voice Training (LSVT) BIG Program demonstrate more of a decrease in motor symptoms than traditional physical therapy. 1

I. A randomized control study that compared LSVT BIG therapy to an exercise program that included only treadmill and seated exercises. It had 11 participants with Parkinson’s disease divided into two groups performing the exercises for 4 weeks. The study looked at the change of motor and nonmotor symptoms.¹

A. Strengths1. The exercise intensity of the General Exercise group were measured by the Borg rating of perceived exertion. This allowed all participants to work at a consistent level instead of some exercises being harder on some patients due to age, condition level, etc. 2. Because the LSVT BIG program requires one-on-one interaction with a PT, this study chose to allow the General Exercise group to also work one-on-one with a PT thus minimize the internal threat of Hawthorne effect.3. This study was double-blinded, minimizing placebo effect and experimenter bias which are both internal threats. 4. Measurements were taken at baseline, 4 weeks, 3 months, and 6 months to see any changed between the groups through time.

B. Weaknesses1. There was a small sample size for this study (11 participants) which leads to an increased risk of type II error, low power, and low effect size.2. All the participants in the study were in the early stage of PD so the more severe effects were not studied in this trial.3. This study excluded participants who had atypical PD, who were already participating in an exercises program, and those with evidence of severe depression. This is an example of the external threat of sample bias which makes the study difficult to generalize to the public4. The General Exercise group combined treadmill training with seated exercise. This is considered multiple treatment interference and is an example of an external threat.

II. A randomized control trial that compared LSVT BIG to a short protocol including multidirectional movements and working on ADLs. The study had 42 patients with Parkinson’s disease who received either 16 sessions of LSVT BIG or 10 sessions of a short protocol. The study compared the effects of these exercises on motor and nonmotor effects.²

A. Strengths1. Examiners found by conducting a power analysis that they needed 17 participants per group and were able to recruit 21 people for each group which decreased the chance of a type II error.


2. Participants were assigned to groups using a random drawing to ensure randomization of groups minimizing internal threats to selection.3. To ensure blinded assessment, the patients were videotaped while performing the UPDRS-III items minimizing experimenter bias.4. During examination, the examiner was blinded not only to group allocation but also to the time the video was taken which decreases experimenter bias.

B. Weaknesses1. Patients for this study were referred by clinics and physicians but the article does not state how the participants were chosen. Without knowing the recruiting method, we are notable to infer the condition of the participants related to those who were not recruited. 2. Participants were excluded if they had dementia, severe depression, disabling dyskinesia, and co-morbidity affectingmobility or ability to exercise. Excluding these participants makes it harder to generalize this study to the public. 3. The two groups received treatment for different lengths of time. The LSVT-BIG group received 16 1-hour sessions (4/week for4 weeks) and the short protocol group received 10 1-hour sessions (5/week for 2 weeks). 4. The post-treatment measurements were taken 16 weeks after the interventions were started. Because the groups haddifferent durations of treatment time, the groups were not being measured at equal times from cessation of treatment.

III. A study comparing LSVT BIG to a walking program and a home exercise program for a change in motor and nonmotor symptoms. The study had 60participants evenly divided into three groups. LSVT BIG and the walking group each had 16 exercise sessions.³

A. Strengths1. The participants chosen for the study fit the inclusion/exclusion criteria as to rule out any alternative explanation for getting better or worse. This decreases a selection threat.2. The study had 60 participants reducing the chance of a type II error.3. This study divided up participants randomly with no differences noted between groups decreasing a selection threat.4. This study was blinded to the rater so to minimize the risk of the internal threat of experimenter bias.

B. Weaknesses


1. In this study, the groups completed their intervention for different amounts of time. The BIG group had 1 hour sessions 4 times a week for 4 weeks while the WALK group had 1 hour sessions 2 times a week for 8 weeks. The article did not state how often the HOME group was to perform their exercises and for what duration.2. The BIG group had one-on-one interactions with the PT while the WALK group had group sessions and the HOME group had no interaction with a PT. This is an example of the Hawthorne effect which is an internal threat. 3. Patients were allowed to perform additional exercises during the study and six participants had changes in their medications during the study. This is an internal threat which could lead to alternative explanations for the results. 4. The WALK group performed their intervention in a park which is not similar to a real clinical situation.

IV. The review of these three RTC leads me to fail to reject the null of my study question. Only one of the studies showed that LSVT BIG improved motor performance better than traditional physical therapy. Of the other two studies, one was unable to detect a difference between the two groups and the other showed that the motor assessment did not differ between the groups.

I also considered the magnitude of effects of the results of the studies when answering my clinical question. The first study¹ had a magnitude of effect of 0.20. It resulted from considerable overlap between data scores and a low sample size. The magnitude of effect of the next study² was 0.41. The last, and most complex study³ had a magnitude of effect of 0.62 between the LSVT BIG group and the WALK group, 0.56 between LSVT BIG and HOME, and 0.11 between WALK and HOME.

These studies failed to show that LSVT BIG has an effect on motor symptoms in patients with Parkinson’s disease (PD). When treating a patient with PD, I would not spend the time and money to either become LSVT BIG certified or find a PT who is since traditional PT shows similar effects. Traditional PT can still have a positive effect on these patients and improve their mobility and other motor symptoms they may be experiencing.

In future studies, more participants are needed with a wider variety of disease levels to be able to generalize the study to the public more easily. Only two interventions should be compared at a time. Lastly, the interventions should be given for the same amount of time and in a one-on-one basis to increase validity and decrease the threats listed in the outline above.


1. Khashayar D, Johnson E, Kani C, et al. Effect of exercise on motor and nonmotor symptoms of Parkinson’s disease. Parkinsons Dis. 2015; 586378.

2. Ebersbach G, Grust U, Ebersbach A, et al. Amplitude-oriented exercises in Parkinson’s disease: a randomized study comparing LSVT-BIG and a short training protocol. J of Neural Trans. 2015; 122:253-256.

3. Ebersbach G, Ebersbach A, Edler D, et al. Comparing exercise in Parkinson’s disease-the Berlin LSVT BIG study. Mov Dis. 2010; 25:1902-1908.


Student ID#: 14147292

Primary Reference: Khashayar D, Johnson E, Kani C, et al. Effect of exercise on motor and nonmotor symptoms of Parkinson’s disease. Parkinsons Dis. 2015; 586378.

Worksheet for Critical Appraisal

(based on PEDro http://www.pedro.org.au/english/tutorial/is-the-trial-valid/ and http://www.ncbi.nlm.nih.gov/pubmed/14703546?dopt=AbstractPlus, which is referred to as a resource by Centre for Evidence Based Medicine )

Instructions. Indicate the correct response by shading it and/or providing an explanation where such a response is requested. If the article did not provide sufficient information for you to be moderately confident about the answer, then add “NS” as an abbreviation for “not stated.” Focus only the methods and results that match your clinical question. You will use and submit this worksheet for Assignments #3 and 5. To make the narrative portion of Assignment #5 easier, please add your own notations inside brackets (e.g., [ ]) to this worksheet. Your notes on this actual worksheet will not be graded. Note. The types of statistical analyses conducted should not to be appraised and the statistical results of a study are neither methodological strengths or weaknesses. Also, “not stated” information is not a weakness for our purposes.

Does the “P” population (patients) of interest in your clinical question match the actual sample of the study? Yes No

Does the “I” intervention in your clinical question match one of the interventions in the study? Yes No

Does the “C” comparison in your clinical question match one of the comparisons in the study? Yes No

Does the “O” outcome measure/s in your clinical question match one or more of the outcome measures in the study? Yes No

Population/patients Is the source from which participants were recruited (e.g., physicians)

described? Yes No Is the method of recruiting (e.g., flyers) participants described? Yes No Is the method of sampling (e.g., convenience) participants described or

discernable? Yes No Is the percentage of participants (i.e., enrolled/consented to participate)

from among the recruits and/or eligible recruits (i.e., met inclusion/exclusion criteria) provided or calculable? Yes No


Were the eligibility criteria (i.e., inclusion/exclusion) appropriate to match your “P”? Yes No

Note. It is very unlikely that a “NS” response is acceptable. Were the eligibility criteria appropriate to reduce alternative explanations for

the results (i.e., extraneous variables)? Yes NoNote. It is very unlikely that a “NS” response is acceptable.

Methodological weaknesses/threats related to “P”: Patients for this study were referred by clinics and physicians but the article does not state how the participants were chosen. Without knowing the recruiting method, we are not able to infer the condition of the participants related to those who were not recruited. This is an example of an internal threat. There was also a small sample size for this study (11 participants) which leads to an increased risk of type II error, low power, and low effect size. Lastly, all the participants in the study were in the early stage of PD so the more severe effects were not studied in this trial.

Methodological strengths related to “P”: The participants chosen for the study fit the inclusion/exclusion criteria as to rule of any alternative explanation for getting better or worse. This is minimizing an internal threat.

Intervention (and Comparison) What research design best matches the design of the study?

Cross sectional (survey) Cross sectional (analytic) Case-control Cohort study Randomized crossover Randomized controlled trial (RCT)

Note. It is very unlikely that a “NS” response is acceptable. Was the comparison group (between)/condition (within) appropriate to

answer the actual research question? For example, real ultrasound compared to sham ultrasound. Yes No

Note. It is very unlikely that a “NS” response is acceptable. Were the participants randomly allocated to groups/conditions? Yes No If the design had groups, were the groups similar (e.g., matched) at baseline?

Yes No NA Were the participants aware of other participants’ groups/conditions?

Yes No Was the intervention /comparison reasonably controlled to reduce

extraneous variables? Yes NoNote. It is very unlikely that a “NS” response is acceptable.


Was the percentage of the enrolled participants who adhered to the intervention/comparison provided (e.g., number of intervention sessions attended) or calculable? Yes No

Was the setting for the study (e.g., community PT clinic) similar to a real clinical situation? Yes No

Methodological weaknesses/threats related to “I/C”: This study excluded participants who had atypical PD, who were already participating in an exercises program, and those with evidence of severe depression. This is an example of the external threat of sample bias and can make the study difficult to generalize to the public.

Methodological strengths related to “I/C”: The exercise intensity of the General Exercise group were measured by the Borg rating of perceived exertion. This allowed all participants to work at a consistent level instead of some exercises being harder on some patients due to age, condition level, etc. Because the LSVT BIG program requires one-on-one interaction with a PT, this study chose to allow the General Exercise group to also work one-on-one with a PT thus minimize the internal threat of Hawthorne effect. Lastly, there was a low mortality rate since all 11 participants completed the trial.

Outcome Measures Were the outcomes measured appropriate to answer the study’s research

question? Yes NoNote. It is very unlikely that a “NS” response is acceptable. Were references for the reliability/validity of the outcome measures

provided? Yes No Were references for the Minimum Clinically Important Difference (MCIDs) of

the outcome measures provided? Yes No Were the outcome measures collected appropriately for reliability and

validity (e.g., bony landmarks used)? Yes No Were the time points for collecting the outcome measures (i.e., testing

intervals and study duration) appropriate to answer the actual research question? Yes No

Note. It is very unlikely that a “NS” response is acceptable. Were the outcome measures collected commonly used in real clinical

situations? Yes NoNote. It is very unlikely that a “NS” response is acceptable. Was the investigator/s, who collected the measures, blinded to participants’

group? Yes No


Was the percentage of participants who completed the outcome measure(s) provided or calculable? Yes No

Describe the magnitude of effects (e.g., differences in changes between groups/conditions in relation to the average variability (SDs) of the groups) regardless of p-value. For example, you might write, “The difference between the means of the two groups at the follow-up measure was about 1/3 of the average SD of the two groups, which seems clinically meaningful based on MCIDs and/or my clinical expertise with the measure.” When comparing the means and SDs of the two groups, an effect size of 0.20 was calculated. This says there was a lot of overlap in the data and the effect size is very small.

Methodological weaknesses/threats related to “O”: The General Exercise group combined treadmill training with seated exercise. This is considered multiple treatment interference and is an example of an external threat.

Methodological strengths related to “O”: The study was double-blinded, minimizing placebo effect and experimenter bias which are both internal threats. Measurements were taken as baseline, 4 weeks, 3 months, and 6 months to see any changed between the groups through time.

(Note. When in doubt about what is “reasonable/appropriate” consider a real clinical situation and/or compare several studies answering similar research questions with similar research designs.)



Primary Reference: Ebersbach G, Grust U, Ebersbach A, et al. Amplitude-oriented exercises in Parkinson’s disease: a randomized study comparing LSVT-BIG and a short training protocol. J of Neural Trans. 2015; 122:253-256.










discernable? Yes No


Is the percentage of participants (i.e., enrolled/consented to participate) from among the recruits and/or eligible recruits (i.e., met inclusion/exclusion criteria) provided or calculable? Yes No


Note. It is very unlikely that a “NS” response is acceptable. Were the eligibility criteria appropriate to reduce alternative explanations for

the results (i.e., extraneous variables)? Yes NoNote. It is very unlikely that a “NS” response is acceptable.

Methodological weaknesses/threats related to “P”: Patients for this study were referred by clinics and physicians but the article does not state how the participants were chosen. Without knowing the recruiting method, we are not able to infer the condition of the participants related to those who were not recruited. This is an example of an internal threat. Participants were excluded if they had dementia, severe depression, disabling dyskinesia, and co-morbidity affecting mobility or ability to exercise. Excluding these participants makes it harder to generalize this study to the public.

Methodological strengths related to “P”. Examiners found by doing a power analysis that they needed 17 participants per group and were able to recruit 21 people for each group which decreased the chance of a type II error.







Yes No


Was the intervention /comparison reasonably controlled to reduce extraneous variables? Yes No

Note. It is very unlikely that a “NS” response is acceptable. Was the percentage of the enrolled participants who adhered to the

intervention/comparison provided (e.g., number of intervention sessions attended) or calculable? Yes No


Methodological weaknesses/threats related to “I/C”: The two groups received treatment for different lengths of time. The LSVT-BIG group received 16 1-hour sessions (4/week for 4 weeks) and the short protocol group received 10 1-hour sessions (5/week for 2 weeks). This could increase the chance of an internal threat.

Methodological strengths related to “I/C”: Participants were assigned to groups using a random drawing to ensure randomization of groups minimizing internal threats. Inclusion/exclusion criteria was also used to minimize internal threats. By making the participants as similar as possible, the researcher decreased the chance of other explanations for changes between groups.


question? Yes NoNote. It is very unlikely that a “NS” response is acceptable. Were references for the reliability/validity of the outcome measures

provided? Yes No Were references for the Minimum Clinically Important Difference (MCIDs) of






group? Yes No


Was the percentage of participants who completed the outcome measure(s) provided or calculable? Yes No

Describe the magnitude of effects (e.g., differences in changes between groups/conditions in relation to the average variability (SDs) of the groups) regardless of p-value. For example, you might write, “The difference between the means of the two groups at the follow-up measure was about 1/3 of the average SD of the two groups, which seems clinically meaningful based on MCIDs and/or my clinical expertise with the measure.” When comparing the means and SDs of the two groups, an effect size of 0.41 was calculated. This is a small effect size.

Methodological weaknesses/threats related to “O”: The post-treatment measurements were taken 16 weeks after the interventions were started. Because the groups had different durations of treatment time, the groups were not being measured at equal times from cessation of treatment. This could be alternative explanations for the results.

Methodological strengths related to “O”: To ensure blinded assessment, the patients were videotaped while performing the UPDRS-III items minimizing experimenter bias. Also during examination, the examiner was blinded not only to group allocation but also to the time the video was taken.




Primary Reference: Ebersbach G, Ebersbach A, Edler D, et al. Comparing exercise in Parkinson’s disease-the Berlin LSVT BIG study. Mov Dis. 2010; 25:1902-1908.










discernable? Yes No Is the percentage of participants (i.e., enrolled/consented to participate)

from among the recruits and/or eligible recruits (i.e., met inclusion/exclusion criteria) provided or calculable? Yes No


Note. It is very unlikely that a “NS” response is acceptable.


Were the eligibility criteria appropriate to reduce alternative explanations for the results (i.e., extraneous variables)? Yes No

Note. It is very unlikely that a “NS” response is acceptable.

Methodological weaknesses/threats related to “P”: Patients for this study were referred by clinics and physicians but the article does not state how the participants were chosen. Without knowing the recruiting method, we are not able to infer the condition of the participants related to those who were not recruited. This is an example of an internal threat.

Methodological strengths related to “P”: The participants chosen for the study fit the inclusion/exclusion criteria as to rule of any alternative explanation for getting better or worse. This is minimizing an internal threat. The study had 60 participants reducing the chance of a type II error.







Yes No Was the intervention /comparison reasonably controlled to reduce

extraneous variables? Yes NoNote. It is very unlikely that a “NS” response is acceptable. Was the percentage of the enrolled participants who adhered to the

intervention/comparison provided (e.g., number of intervention sessions attended) or calculable? Yes No



Methodological weaknesses/threats related to “I/C”: In this study, the groups completed their intervention for different amounts of time. The BIG group had 1 hour sessions 4 times a week for 4 weeks while the WALK group had 1 hour sessions 2 times a week for 8 weeks. The article did not state how often the HOME group was to perform their exercises and for what duration. Also, the BIG group had one-on-one interactions with the PT while the WALK group had group sessions and the HOME group had no interaction with a PT. This is an example of the Hawthorne effect which is an internal threat. Also, patients were allowed to perform additional exercises during the study and six participants had changes in their medications during the study. This is an internal threat which could lead to alternative explanations for the results. Lastly, the WALK group performed their intervention in a park which is not similar to a real clinical situation.

Methodological strengths related to “I/C”: This study divided up participants randomly with no differences noted between groups. This minimized the risk for internal threats such as statistical regression and selection.


question? Yes NoNote. It is very unlikely that a “NS” response is acceptable.

Were references for the reliability/validity of the outcome measures provided?

Yes No Were references for the Minimum Clinically Important Difference (MCIDs) of






group? Yes No Was the percentage of participants who completed the outcome measure(s)

provided or calculable? Yes No


Describe the magnitude of effects (e.g., differences in changes between groups/conditions in relation to the average variability (SDs) of the groups) regardless of p-value. For example, you might write, “The difference between the means of the two groups at the follow-up measure was about 1/3 of the average SD of the two groups, which seems clinically meaningful based on MCIDs and/or my clinical expertise with the measure.” When comparing the means and the SDs of LSVT BIG and WALK, an effect size of 0.62 was found. This is considered a medium/big effect size. When comparing LSVT BIG and HOME, an effect size of 0.56 was found. Lastly, when WALK and HOME were compared, an effect size of 0.11 was found which is considered a small effect size. These were calculated based only on the UPDRS III which was the main measure of the study.

Methodological weaknesses/threats related to “O”: The post-treatment measurements were taken 16 weeks after the interventions were started. Because the groups had different durations of treatment time, the groups were not being measured at equal times from cessation of treatment. This could be alternative explanations for the results. Lastly, one of the result measurements was the timed up-and-go. Because this is a walking task, the participants in the WALK group had more indirect practice with this skill than the other 2 groups. This is an example of the internal threat of testing.

Methodological strengths related to “O”: This study was blinded to the rater so to minimize the risk of the internal threat of experimenter bias.


Documents

Assignment 4