Venous Ulcers, Falls, Balance Confidence, and Strength: Cause and Effect

    loading  Checking for direct PDF access through Ovid

Excerpt

This month’s continuing education article, “Falls, Balance Confidence, and Lower Body Strength in Patients Seeking Outpatient Venous Ulcer Wound Care” is a prime example of the need to establish cause and effect1 when attempting to assign a given medical condition, comorbidity, or wound type to its effect on physical function. As the authors and others note, the etiology of falls is multifactorial,2-5 as are the causative factors of dysfunctional balance and weakness of the skeletal, truncal, and core muscles. The typology of lower-extremity ulcers is notwithstanding.
How then do we establish cause and effect4 when considering the relationship of venous ulcers (VUs) to the cause or influence on dysfunctional balance and falls? If the authors refer to this study as a “quality improvement project,” cause and effect should be established. There are several ways to accomplish this; one is through the internal validity of the study.
The concept of internal validity refers to how well a research methodology or the design of the protocol works to prevent the influence of confounding variables (>1 possible independent variable or cause acting at the same time) in a study.4,6 A high level of internal validity gives researchers the ability to test a hypothesis.6 A research study with suitable internal validity lets the investigator choose 1 explanation over another with confidence because it mitigates (many possible) confounding variables,6 that is, a central or a peripheral central nervous lesion or such conditions that have significance influence on balance, mobility, and falls but have no relationship to VUs. When it comes to treatment interventions, we must ask, “Is the relationship causal between the treatment and the outcomes, and what influence do the independent and dependent variables (alternative causes) have on the observed effect?” External validity refers to the generalizability of the research that has the ability of its conclusions to be valid and extend from the specific laboratory or research environment to a similar “real world” cause-and-effect situations (from statistical significance to clinical relevance).6
There are 3 main criteria for establishing cause and effect: (1) Temporal precedence (cause then effect over time); which came first: the weakness, fall, or the drug-related VU or the venous hypertension ulcer? (2) Is termed “covariation” of the cause and effect. For example, if X and then not Y and, in contradistinction, if not X then not Y (if an intervention is given, then outcome is observed [usually if intervention is not given, the outcome is not observed]). (3) If there are no alternative explanations for the weakness and falls, a treatment can be instituted; this is termed micromediation—the intervention is balance and strength training, and a specific outcome measured—to increase strength and decrease falls. If the desired outcome is not achieved, there may be alternative causes termed substantive or nuisance. The central focus of applying internal validity is to “rule out” alternative explanations. Experiments can test causal claims and be strengthened by the random assignment that also controls for possible alternative explanations, thus strengthening the research. However, a special caveat about validity: Increased internal validity has the consequence of potentially decreasing external validity, which restricts generalizability or clinical significance of a study or narrows the focus. Experiments are studies designed to show causality, which depends on internal validity. It has been said that there is tension between emphasizing causality and the generalizability.4 The randomization of participants and random assignment of treatments are the most powerful ways of controlling controlling threats to internal and external validity.4,6
In the continuing education article on page 85, the authors used a cross-sectional methodology. Because the analyses presented are cross-sectional, cause-and-effect relationships cannot be proven.
    loading  Loading Related Articles