Quantitative Research Critique: Evaluating the Impact of a ...



Quantitative Research Critique: Evaluating the Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults – A Nonequivalent Control Group Design

Tonya Allendorfer, Melissa Hayes, Chelsea Melrose, Kathleen Michalak

Ferris State University

Abstract

The purpose of this assignment is to demonstrate how to critique a nursing research study. The study to be critiqued was conducted at The Miriam Hospital which is a 247-bed private, not-for-profit, acute care teaching hospital in Providence, Rhode Island. The aim of the study was to determine the impact of a nurse-driven mobility protocol on maintaining functional status. At the conclusion of the study, it was determined that early and ongoing ambulation may be an important contributor to shortening length of stay and preventing functional decline. This critique will include an evaluation of the study’s objectives, significance, sources, theoretical framework, hypothesis, data collection methods, and instrumental validity and reliability. Study data, results and conclusions along with limitations to recommendations from researchers, will also be analyzed throughout this critique.

Quantitative Research Critique: Evaluating the Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults – A Nonequivalent Control Group Design

Purpose and Problem Statement

Identifying the purpose and problem statement is the first step of the critique process. The problem statement should be evaluated for clarity. It should be written as a single declarative or interrogatory sentence (Nieswiadomy, 2012). Study variables and the population to be studied should be described in this statement. Study feasibility, ethical aspects, and the significance of the study to nursing should also be reflected in this statement.

Purpose

Evidence. In reviewing the article, Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults, there was a definite purpose. The purpose was stated right away in the abstract of the article.

“The purpose of the study was to determine the impact of a nurse-driven mobility protocol on functional decline. A nonequivalent control group design was used; the independent variable was mobility protocol and dependent variables were functional status and length of stay. Older adults who participated in a mobility protocol maintained or improved functional status and had a reduced length of stay. Practice implications include an emphasis on ambulation in hospitalized older adults” (Padula, Hughes & Baumhover, 2009, p. 325).

This is a relevant area of study because older adults often become debilitated in the hospital.

Support. A good purpose statement clearly sets up the problem of the study. According to Burns & Groves, “The purpose also includes the variables, the population, and often the setting for the study” (p. 146). You can often get a sense of the purpose from the title of the article. The purpose statement is stated in the abstract of the article most of the time and at the very least on the first page. It should also be reiterated after the literature review (Burns & Groves). Niewsiadomy’s critique guidelines are a little more specific. These guidelines suggest the purpose statement be in one sentence. In this one sentence, it wants to know the significance, feasibility, data, and if the study is ethical. The purpose statement contains a lot of information regarding the content of the study that a potential reader would need to know.

Analysis. In the article, Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults, purpose is strongly supported. Although it was not in a single sentence it did clearly state the problem, which was to “determine the impact of a nurse-driven mobility protocol on functional decline” (Padula, Hughes & Baumhover, 2009, p. 325). The variables in this study included the mobility protocol, functional status, and length of stay. The hospitalized older adult is the population and setting. The article did give a sense of the data that would be learned by stating the participants stayed the same or improved, but did not provide specifics. It was an ethical study to improve the lives of hospitalized patients and no treatment was withheld. The significance and feasibility were not evident in the purpose statement, but were provided later in the article. The article was directed at nurses and might be difficult for people without a medical background to understand. Overall, the purpose statement was supported.

Problem Statement

Evidence. After reviewing the study written by Padula, Hughes & Baumhover the problem statement was discussed within the first two paragraphs of the article. The authors presented it by starting out discussing what functional status is and why it is important. Functional status is the ability to perform basic self-care activities and is a significant component in the health and quality of life of older adults. Functional status must be maintained in order for older adults to continue to be independent and healthy (Padula, Hughes & Baumhover, 2009). A component of functional status is mobility. Maintaining mobility is vital in preserving independence in activities of daily living such as bathing, getting dressed and feeding ones’ self. After the authors defined what functional status is and established its importance in the older adult population, the problem was defined. Maintaining functional status is important; however research has demonstrated that low mobility and bed rest are common during acute hospitalization. Research shows that older adults who have an acute hospital stay often result in complications that lead to a decline in their functional status.

Support. According to Burns and Groves, a problem statement in a study should include the following:

1. Identifies an area of concern for a particular population

2. Indicates the significance of a problem

3. Provides background for the problem

4. Outlines the need for additional study in a problem statement (Burns & Groves, 2011).

According to Burns & Groves, the study in review has a problem statement that incorporates all of these components. However, the example in the book shows that a problem statement is just that…a statement. The problem in the study being reviewed clearly defines a problem; however it is over two paragraphs. It is not easily picked out as in the text examples given by Burns & Groves.

Analysis. The evidence in the research article clearly defines a strong problem area in the decline in functional status of older adults during an acute hospitalization. The problem is feasible and ethical and pertains to the nursing and health professions. The problem is mixed in throughout the study, not clearly defined in one problem statement. By not having a clear problem statement, it is hard for the reader to note why the authors felt this problem was or is important. If someone without a background in healthcare were to read this article the problem would not easily be picked up on until reaching the purpose section, which is several paragraphs into the article.

Review of Literature

Evidence. Evidence as provided by the mobility article written by Padula, Hughes & Baumhover indicates that older adults experiencing an acute hospital stay often times suffer a functional decline (Brown, Friedkin, & Inouye, 2004; Covinsky, Palmer, & Fortinsky, 2003; Inouye, Wagner, Acampora, Horwitz, Cooney, & Tinetti, 1993; Brown, Roth, Peel, & Allman, 2006).

Support. A prospective multi-center study was used to examine the effects of low mobility and bed rest in patients older than 70 years of age. The results of this study determined that functional decline may be a result of iatrogenic events (Brown et al., 2004). A worsening of ADL function at discharge as compared with preadmission activity levels was determined in a descriptive study of 2,293 older adults. This happening resulted in thirty-five percent of patients studied during an acute hospital stay (Covinsky et al., 2003). Patients experiencing functional deficits as a result of hospitalization are less likely to recover lost function. This finding was a result of a study involving 1,270 older adults who were hospitalized for acute illnesses (Inouye et al., 1993). Mobility was found to be related to recovery of ambulatory ability in a study of 285 patients who were expected to be limited in mobility for the first five days of hospitalization (Brown et al., 2006).

Analysis. The mobility article as written by Padula, Hughes & Baumhover provides little in the way of a literature review. The sources of information used for this study appear to be secondary sources. The sources were published in years prior to the study and were not developed as a part of the study. Some of the studies discussed were greater than five years old and were summarized rather than critiqued. The references appear to be free of citation errors, but are not necessarily in APA format. As a nurse, the reader may be able to determine the importance and relevance of the study findings discussed. However, all disciplines in practice may not be able to gain insight from this article.

Theoretical/Conceptual Framework

Evidence. Evidence as provided by the mobility article written by Padula, Hughes & Baumhover indicates that the healing process can be improved by getting patients out of bed and moving around earlier than usual (2009). The Geriatric Friendly Environment through Nursing Evaluation and Specific Interventions is implementing this theory for Successful Healing (GENESIS) (Burns & Grove, 2011). The whole mobility article is based off the Self-Care Deficit Theory of Nursing provided in the Understanding Nursing Research book written by Burns & Grove.

Support. The GENESIS program was used in this mobility article as a way to implement the whole theory of the paper. The GENESIS program was used as a model of nursing care delivery in everyday operations. The program included an extensive three-day continuing education for all geriatric nurses in all units. This program includes the nurse-driven mobility protocol which served as the intervention for this research. As part of this program, all registered nurses were required to question all orders for bed rest and estimate needs and hurdles to mobility. It is the priority of all nurses to get patients up and out of bed, who are a part of the GENESIS program and walk with them three to four times a day. Nurses are also required to assist patients to a chair for meals and help patients to the bathroom (Burns & Grove, 2011). This mobility article was based on the Self-Care Deficit Theory of Nursing found in Understanding Nursing Research by Burns & Grove, 2011. The goal of this theory is to have patients/individuals perform and start self-care activities in order to maintain activities of daily living and increase personal growth (Padula, Hughes & Baumhover, 2009).

Analysis. The mobility article written by Padula, Hughes & Baumhover makes it difficult to find the conceptual framework (CF). This is a weak section in the article. A clearly defined CF was not found. The article talks about GENESIS, but does not provide a clear definition of what this program is about. There were no clearly defined terms, assumptions, or propositions as defined in the Nieswiadomy critique guidelines. The Self-Deficit Theory of Nursing provides a good description of what the definition is. It is clear to any reader that this mobility study is based on this theory.

Hypothesis

Evidence. The first hypothesis states that older adults who participate in a mobility protocol will maintain or improve functional status from admission to discharge. The second hypothesis states older adults who participate in a mobility protocol will have a reduced length of stay.

Support. A hypothesis is a formal statement of the expected relationship(s) between two or more variables in a specified population (Burns & Groves, 2011). The hypothesis translates the research problem and purpose into a clear explanation or prediction of the expected results or outcomes of selected quantitative and outcome studies (Burns & Grove, 2011). Hypotheses are broken into different types:

- Simple hypothesis: states the relationship (associate or causal) between two variables.

- Complex hypothesis: states the relationships (associate or causal) among three or more variables.

- Nondirectional hypothesis: states that a relationship exists but does not predict the nature of the relationship.

- Directional hypothesis: states the nature (positive or negative) of the interaction between two or more variables.

- Null hypothesis: is used for statistical testing and for interpreting statistical outcomes. It is used when the researcher believes there is no relationship between two variables and when theoretical or empirical information is inadequate to state a research hypothesis.

- Research hypothesis: is the alternative hypothesis to the null hypothesis and states that a relationship exists between two or more variables.

Analysis. In the mobility article the two hypotheses analyzed are clearly stated, early in the article and also in the discussion section at the end. The authors of the article do not state the null hypothesis in the article, however it is implied because the null hypothesis is opposite of the research hypotheses stated. One can assume the null hypotheses for this article would be that there would be no difference in the functional status of patients involved in the mobility protocol and those not involved, nor will there be a difference in length of stay between those in the mobility protocol and those who are not involved.

This study clearly states the hypotheses and it is clearly worded and concise. The hypothesis is clearly tied to the study problem, which is the functional decline of elderly patients admitted to the hospital.

Research Design

Evidence. In the article, Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults, the study was stated as using a nonequivalent control group research design.

Support. According to Burns & Grove (2011), experimental research has a tight control of the study by random assignment of subjects to groups and the researcher’s ability to manipulate the independent variables. When complete control is not possible, researchers use quasi experimental designs with comparison groups. In nonequivalent control group designs, there is only one group. Only having one group weakens the validity. To strengthen the validity of a study, two groups would be needed that could be tested pre intervention and post intervention.

Analysis. In the article, Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults, research design is weak. There are too many uncontrolled variables that interfere with the validity. The treatment group had 5.43 RN hours per day, while the control group only had 5.32. There are many studies showing that increased RN hours directly correlates with better patient outcomes. Another element that weakens the study is that they choose medical patients. That is very broad. It could have narrowed by choosing only pneumonia patients that were being treated with the same protocol. The fact that the control group was the only unit left in the hospital that had not gone to the GENESIS protocol, leads the reader to think that the researcher might have some knowledge of what was being done with the treatment group.

Population, Sample & Sampling Plan

Evidence. In the article, Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults, a convenience sample of adults aged 60 years or older; who were admitted with medical diagnoses to one of two nursing units was used for this study. In order to meet study criteria, the patients must have had a length of stay of three or more days and an ability to understand English. They must also have been without a physical impairment that would significantly limit ability to mobilize and must have been cognitively intact or with a significant other who was able to participate. A Mini Mental Status Exam was used to determine cognitive status. A score of 24 or more was required to be included in the study. Because there are potential limitations in mobilization associated with the surgical experience, medical patients were used. A total of 453 patient records were screened for eligibility. 84 eligible subjects were enrolled. 34 subjects were withdrawn from the study based on criteria requirements. Study data were collected within 48 hours of admission. Sample size statistics determined adequate power and supported a total number equal to 50 (Padula et al., 2009, p. 327).

Support. The researcher randomly selects and assigns members to groups through probability sampling. “In probability sampling, every member has a probability higher than zero of being selected for the sample. To achieve this probability, the sample is obtained randomly” (Burns & Grove, 2011, pp. 298-299). Four sampling designs are used in probability sampling: simple random sampling, stratified random sampling, cluster sampling, and systematic sampling.

An example of simple random sampling is taking the names of individuals meeting criteria and placing them in a hat. After the names are in the hat, each one is randomly selected until the desired number of participants is achieved. Stratified random sampling occurs when the researcher uses the variables of the sample to stratify those who are represented. For example, the researcher may include a certain number of individuals from each ethnic background or in differing age ranges meeting criteria. Variables used for stratification must be defined in the research report. Cluster sampling occurs when “a researcher develops a sampling frame that includes a list of all the states, cities, institutions, or organizations with which elements of the identified population can be linked” (Burns & Grove, 2011, p. 302). A randomized sample of these elements can then be used in the study. When an ordered list of all members of the population is available, systematic sampling may be used. Systematic sampling “involves selecting every kth individual on the list, using a starting point selected randomly” (Burns & Grove, 2011, p. 303).

Without randomly selected samples, internal validity or accuracy of the results may be compromised. The researcher must be able to generalize findings to the entire population in order for external validity to occur. That is, the represented sample must not be closed to a certain ethnicity or age group, but rather reflective of all age ranges and ethnicities meeting criteria.

Analysis. In order to ensure internal validity, the researcher must identify the population from which the sample will be taken and then randomly select the sample to be included in the study. Those included in the population which will be randomly selected must be similar in characteristics or variables which make them eligible for study. A written study report must outline these variables and must include how the sample was selected and by whom. Without each of these areas being clearly defined, there is question as to whether the results are due to our treatment or due to variables between groups.

There are many issues with the sampling from this article. It does identify the patients that are targeted, but this could be more specific. The study uses patients from two separate units in a magnet status hospital. It is not discussed what would happen if the facility was not already magnet status or what the differences in the units are. This affects external validity. The study would have more validity if the researchers would have chosen only pneumonia patients. This was a small sample of fifty patients from one hospital in which the rest of the hospital had already gone to this particular protocol that was being tested. The researchers discuss bias as being similar in units, but the treatment group had 7.18% nursing hours versus a control group with 6.91% nursing hours. It is assumed that those nurses not getting patients up would state that this was due to a lack of help or time as the top reason for not doing so. This small percentage makes a huge difference. The researchers do a good job discussing why patients dropped out of the study. Most patients that were removed either moved or were discharged before the three day mark.

Levels of Measurement and Data Collection

Evidence. In the article, Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults, the only level of measurement found included the measurement of cognitive statuses by the completion of a Mini-Mental Status Examination with a score of 24 or more. The data was collected within 48 hours of admission for eligible subjects. For the GENESIS program, nurses were the main source of data collection through a series of questions. All data was collected at the Miriam Hospital in Providence, Rhode Island.

Support. Levels of measurement are rules which were developed in 1946 for assigning numbers to objects in order for a hierarchy in measurement to be established. There are four measurement categories. Nominal-scale measurement is the lowest of the categories. It is used when data can be organized into categories of a defined property but the categories cannot be order ranked. Ordinal-scale measurement is data that is assigned to categories that can be ranked. Interval-scale measurement uses scales which have equal numerical distances between intervals. Ratio-scale measurement is the highest form of measurement and meets all of the rules for other forms of measurement. According to Burns & Grove (2011), there are five tasks to be completed when collecting data. These tasks include: selecting subjects, collecting data in a consistent way, maintaining research controls as indicated in the study design, protecting the integrity of the study, and solving problems that threaten to disrupt the study.

Analysis. This section of the study is weak. The authors do a poor job in explaining who, what, when, where, and why of data collection. The instruments and the levels of measurement for the data being collected are unclear. The article states that SigmaStat, descriptive statistics and inferential statistics were used, but it does not discuss how this affected the study. It simply talks of ensuring adequate power and variables. There is no discussion of what the variables are or what the differences between the groups are. Without this discussion, one can not assume that the statistical tests which were used were correct for this study.

Instruments

Evidence. This study utilized a demographic data collection sheet which was developed specifically for the research in the article. This included the following information:

- Age

- Gender

- Primary diagnosis

- Use of assistive devices

- Fall-risk assessment

- Presence of any restrictions to mobility

- Use of PT/OT

- Type of out-of-bed activity

Functional status was configured by using the following:

- The modified Barthel Index (BI): a subjective measure that measures the capacity of an individual to perform identified Activities of Daily Living (ADL) such as personal hygiene, bathing, feeding, toileting, etc.). The modified version used 10 items and included a 5 point rating scale for each item. A quantitative estimate of the level of dependence was obtained, ranging from 0 (totally dependent) to 100 (totally independent). The BI was used to assess self-perceived level of function at the time of admission, for two weeks prior to admission and at discharge.

- Up and Go Test: objective assessment that measures subjects’ ability to stand from an armchair, walk three meters, turn around, return and sit down in the chair again. Researchers also used a “mini-mental state exam” to evaluate cognitive status.

Researchers proposed that the BI was the standard of care for clinical and research purposes. The researchers also state that the Get Up and Go test has been shown to be “reliable and valid” and is reported to correlate with the BI at r = -0.78.

Support. In research, it is very important to have trustworthy data; especially if a study is to produce useful findings to guide nursing practice. Quality measurement methods are used in a study to help improve the accuracy or validity of study outcomes or findings.

Validity:

The validity of an instrument is a determination of how well the instrument reflects the abstract concept being examined. Validity is not an all-or-nothing phenomenon; it is measured on a continuum. No instrument is completely valid. Validity varies from one sample to another.

- Construct validity: a single broad method of measurement evaluation which includes content and predictive validity.

- Content-related validity: examines the extent to which the measurement includes all the major elements relevant to the construct being measured.

Three common types of validity presented in published studies include evidence of validity from:

- Contrasting groups: indentifying groups that are expected (or known) to have contrasting scores on the instrument.

- Convergence: determined when a relatively new instrument is compared with an existing instrument that measures the same construct.

- Divergence: study participants complete both of these scales to examine evidence of validity from divergence.

Reliability is concerned with the consistency of the measurement method. Reliability testing is a measure of the amount of random error in the measurement technique.

Analysis. When discussing the instruments used in this research article, the researchers do a good job of explaining the tests that were used to examine functional status of the participants. However, when it comes to discussing the validity of these tests, it appears to be somewhat lacking. The researchers included how the instruments correlated with each other, but did not present the internal or external validity of their instruments. No pilot study was done prior to this study.

Descriptive Statistics

Evidence. The article Impact of a Nurse-Driven Mobility Protocol on Functional Decline in Hospitalized Older Adults states in the data analysis section “Descriptive statistics were completed for study variables and comparisons between the groups on baseline characteristics were examined” (p. 329). The results section of the article describes the sample as being 23 men and 27 women between the ages of 62 and 97. The study states 40 patients lived at home, 7 in assisted living, and 3 were admitted from a nursing home with 31 of them having to use some sort of assistive device. Out of the 50 people, 20 had a physical therapy consult and 8 had an occupational therapy consult. The study does mention that there were no significant differences in demographic variables.

Support. Understanding Nursing Research by Burns & Grove states descriptive statistics “includes frequency distributions, measures of central tendency, measures of dispersion, and standardized scores” (p. 383). Frequency distributions are a way to organize measures of a variable and are either grouped or ungrouped. A measure of central tendency is the average of the collected data. Measures of dispersion are used with measures of central tendency. Dispersion is how the individual sample compared with the average. Standardized scores are sometimes used to do this. Some things are difficult to compare so they “transform raw scores into standardized scores” (p. 388). These are specific to each study.

Analysis. Impact of a Nurse-Driven Mobility Protocol used a table to represent nursing staff characteristics to compare the two units of the study. The text states that these units are similar and comparable, but this author disagrees. The treatment group received more RN and unlicensed assistive personnel hours than the control group. The treatment group received 7.18% total nursing hours, while the control group only received 6.91% (p. 327). The only other table the study listed was comparing the Barthel scores. The Barthel score was the patient’s perceived level of functioning. Again, the treatment group was at an advantage because their scores were higher to begin the study with. The study did use some central tendency stating the mean age was 80.4 years. A mean number of diagnoses of 6.7 were also used, which is interesting because the only other mention of diagnosis is the fact that the subjects had to be medical patients. The article did not state if the previously mentioned physical and occupational therapy consults were included in the treatment or control group or if they made a difference in the study. Overall, this was not a really strong portion of the study. Details were lacking.

Inferential Statistics

Evidence. According to Baumhover, Hughes & Padula, both of the hypotheses were supported by descriptive and inferential statistics. “Hypothesis 1, that older adults who participate in a mobility protocol will maintain or improve functional status from admission to discharge was supported (P=0.05). Hypothesis 2, older adults who participate in a mobility protocol will have a reduced LOS, was also supported (P ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download