dpph seminar: possibilities of various study designs, lammi sept. 28-29, 09

DPPH Seminar: Possibilities of various study designs, Lammi Sept. 28-29, 09

Group work: Critical appraisal

Jani Ruotsalainen, MSc, BSc, etc.

Understanding OH intervention studies / FIOH / Jani Ruotsalainen / 21.04.23 2

Overview of group work

1) Warm-up by coding abstracts with the COHF system• Discuss correct answers

2) Read article by Jensen and Friche 2008• Answer questions• Discuss results

Time permitting...

3) Read article by Steenstra et al. 2006• Answer questions• Discuss results


Exercise 1

• Code the abstracts you are given according to:

Study designsA1 Randomised Controlled Trial (RCT)

A2 Controlled Before-After study (CBA)

A3 Interrupted Time-Series (ITS)

A4 Before-After study (BA)

OH outcomes (NB. what was measured?!)B1 Exposure (to e.g. chemicals, noise, stress)

B2 Behaviour (e.g. wearing protective equipment)

B3 Occ. disease and symptoms (e.g. eczema, asthma)

B4 Disability, sickness absence, return to work

B5 Injuries

B6 Quality of OH services

B7 Public health at the workplace

One of these

One or more of these


For example

Ketola, R. et al. Effects of ergonomic intervention in work with video display units SJWEH, 28[1], 18-24. 2002.

OBJECTIVES: This study evaluated the effect of an intensive ergonomic approach and education on workstation changes and musculoskeletal disorders among workers who used a video display unit (VDU). METHODS: A randomized controlled design was used. The subjects (N=124) were allocated into three groups (intensive ergonomics, ergonomic education, reference) using stratified random sampling. The evaluation involved questionnaires, a diary of discomfort, measurements of workload, and an ergonomic rating of the workstations. The assessments were made 2 weeks before the intervention and after 2 and 10 months of follow-up. RESULTS: The intensive and training groups showed less musculoskeletal discomfort than the reference group after 2 months of follow-up. Positive effects on discomfort were seen primarily for the shoulder, neck, and upper back areas. No significant differences were found for the strain levels or prevalence of pain. After the intervention the ergonomic level was distinctly higher in the intensive ergonomic group than in the education or reference group. CONCLUSIONS: Both the intensive ergonomics approach and education in ergonomics help reduce discomfort in VDU work. In attempts to improve the physical ergonomics of VDU workstations, the best result will be achieved with cooperative planning in which both workers and practitioners are actively involved.

Coding: A1, B1, B3


Small group exercise

• You all have a list of abstracts

• Now, in pairs or in groups of three read through the six abstracts and code each one according to the COHF system (I will leave it visible on here)

• After 15 minutes we discuss


Correct answers

Bøggild H, Jeppesen HJ. Intervention in shift scheduling and changes in biomarkers of heart disease in hospital wards. A2, B1, B3

Rasmussen K, et al. Prevention of farm injuries in Denmark. A1, B2, B5

Smits PB, et al. Problem-based learning versus lecture-based learning in postgraduate medical education. A1, B6

Wergeland EL, et al. A shorter workday as a means of reducing the occurrence of musculoskeletal disorders. A4, B1, B3

Hanlon P, et al. Health checks and coronary risk: further evidence from a randomised controlled trial. A1, B7

Joy GJ, Middendorf PJ. Noise exposure and hearing conservation in U.S. coal mines--a surveillance report. A3, B1, B2


Exercise 2: Jensen & Friche article

Scenario: Imagine you are an occupational physician. You know that knee problemsare common in the construction industry. You consider organising an expensive trainingregime to decrease knee problems. You wonder if there is evidence that such traininginterventions actually do decrease knee problems. Whilst searching PubMed you find theJensen & Friche article.

Read the article and answer the following questions:

1. How would you formulate a research question that would answer your problem from practice? (use PICO)

2. What is the research question of the study according to the title?

3. What is the research question that is relevant to your problem from practice?

4. For which intervention and for what outcome do the authors present their results in the abstract?

5. What were: a) the magnitude of the outcome and b) the uncertainty related to the above?

6. Do these results give an answer to the original research question and the question you have from practice?

7. Where are the answers to the original research questions presented in this article?

8. Can you calculate a risk difference or a risk ratio that would answer your question from practice?

9. Should you implement the intervention in your practice?


Research questions

• What are the research questions?• Do the results presented answer the research questions?

1. Was the frequency of floor layers using the new working methods still the same after a period of two years?

2. Did the use of new working methods spread to floor layers who did not participate in courses?

3. Did the use of the new working methods reduce MSK complaints in the knee when comparing floor layers who changed their working methods with those who did not?

4. Did the use of the new working methods result in other health problems: MSK complaints from elbows, wrists and back?

5. In the data of the 2003 study, could predictors be found of the presence or absence in 2005 of sustained more serious knee problems?


Results as comparisons

What is interesting:

Comparison Outcome

training in new work methods -> knee complaints (less?)vs

no training in new work methods -> knee complaints

What was reported:

Comparison Outcome

use new methods <1 year -> knee complaints >30 daysvs

use new methods ≥1 year -> knee complaints >30 days

(results adjusted for age, BMI and stress)


Let's take a closer look at the results

• Using new working methods might prevent moderate-severe knee pain

• Using new working methods when you have knee complaints might decrease them

• Reported as risk increase if NOT using new methods

Let's redo the calculations

28

59

<30 days

(6/65)/(5/33)

5

6

>30 days

Outcome: Develop knee complaints

33Use methods <1year

= 0,61Risk ratio =

65Use methods >1year

TotalIntervention

28

59

<30 days

(6/65)/(5/33)

5

6

>30 days

Outcome: Develop knee complaints

33Use methods <1year

= 0,61Risk ratio =

65Use methods >1year

TotalIntervention


Can we find what is interesting?

• Table 2 about halfway

GroupsUse new working method Training Control

Never 23 139

Occasionally 59 83

Weekly 41 32

Daily 10 9

Sum of knee complaints 133 263

Total at risk 216 454

Risk133/216 =

0,62263/454 =

0,58


Exercise 3: Steenstra et al. article

Scenario: As an occupational physician you see many workers who have low back pain with associated disability and absence from work. You wonder what are effective interventions in increasing return to work (RTW) in patients with these complaints. After a small search in PubMed you stumble upon the article by Steenstra et al. on the effectiveness of graded activity. You try to infer from the results in the article if this is something that you should implement in your own practice.

Read the article and answer the questions on the exercise sheet


Thank you for your attention!

dpph seminar: possibilities of various study designs, lammi sept. 28-29, 09

Documents