TITLE: Article Analysis



TITLE: Article AnalysisSUMMARY:A common uncomfortable condition called postoperative ileus (POI) frequently occurs after abdominal surgery. There is not an effective treatment for it and it is uncomfortable for the patient and frustrating for the health care provider who is unable to help resolve the condition. The research article “A Randomized Trial of Rocking-Chair Motion on the Effect of Postoperative Ileus Duration in Patients with Cancer Recovering from Abdominal Surgery,” addresses that issue.It is thought that a POI is the result of the body’s sympathetic response to the surgical stress, the manipulation of the bowel, and abdominal incisions. Bowel sounds are absent and gastrointestinal function slowed. A buildup of gas that cannot be expelled becomes extremely painful causing abdominal distention, nausea, and vomiting. A POI can last up to 7 days post-operative and other than reassure the patient it will improve; health care providers have little else to offer the patient in treatment. A common accepted practice for post-operative patients in hope of resolving POI is to get the patient moving on post op day 1. Patients are encouraged to get out of bed, sit in a chair, and walk. There is no convincing evidence that this practice is effective for POI. Previous research has found those interventions to be effective at preventing other post op complications. An intervention that has shown promise in reducing pain medication use in women who have had an abdominal hysterectomy or cesarean section is rocking in a rocking chair. Those studies have also indicated that patients who rock in a rocking chair post op passed flatus earlier and discharged home sooner. The study completed had inclusion criteria of age 21 or older, scheduled to undergo abdominal surgery for gastrointestinal cancers, and scheduled to have a patient-controlled-analgesia (PCA) pump. The ability to read, write and understand English was also criteria. The final criteria for inclusion was that the patient be cognitively intact and be able to tolerate sitting and rocking in a chair.The study was conducted on a 32 bed oncology unit. Statistical software, SPSS Sample Power 2.0, was used to set significance at .05, two-tailed tests, and a power of .80. Sample size needed to be at least 54. The three research questions were evaluated. “Does the rocking intervention reduce the mean time in days to passage of first flatus” (Massey, 2010)? “Does the rocking intervention reduce the total mean pain Morphine Equivalent Dose (MED) medication in milligrams received compared to standard care” (Masssey, 2010)?“Does the rocking intervention reduce the mean time in days to the hospital discharge compared to standard postoperative care” (Massey, 2010)?The study was a posttest-only randomized control trial design and had 66 patients (n=32 non-rocking and n=34 rocking). The investigator met with each participant daily until the first passage of flatus. Other data collected included demographic information, diagnosis and surgical procedure, anesthesia and surgical time, and history of previous abdominal surgeries. Pain medication usage was retrieved from the infusion pump daily, and discharge information retrieved from the hospital computer system. The article does not specify how the data was obtained on laps ambulated or hours in rocking or non-rocking chair.Study results found that rocking had significantly reduced the time to the first flatus but was not statistically significant for a reduction in pain medication use and a reduction of time spent in the hospital. The following statistics were used in the research article.Name of the statistic-MeanWas this statistic covered during the class? Yes, it is a measure of central tendencyNumber of times used in article: 9Why do you think this statistic was used? The mean is the most frequently used measure of central tendency and is best used with interval data, meaning data that does not have a natural starting point. The mean gives an “average” of a set of numbers. The mean was used in the article to give an “average” for the rocking and non-rocking in the following areas:Surgical characteristicsSubject characteristics-gender, ethnicity, marital status, and diagnosis.[Median was used for age.]Surgery durationAnesthesia durationTimes to first flatusTotal pain medication receivedTime in rocking/non-rocking chairsTime to dischargeLaps ambulated per dayThe statistic was necessary to the study to give a measure of center (average) to the various areas it was used in.What did this statistic show or prove: The mean showed a measure of center of the data.Was there an associated P value? There was a P value associated with the data categories including the mean. P values are discussed in the standard deviation and t-test sections.If yes, what did the P value show? See standard deviation and t-test sections2. Name of the statistic: MedianWas this statistic covered during the class? YesNumber of times used in article: 1Why do you think this statistic was used? The median is a measurement of central tendency. The median is frequently used by epidemiologist for age. The median is not as sensitive to extremely high or low data entries. The median was used in this study only in regards to age, the mean was used as the measurement of central tendency for all other areas. What did this statistic show or prove: The median shows the “average” age of study participants.Was there an associated P value? YesIf yes, what did the P value show? The P value is P=.600. That is statistically not significant, which means that there was not a significant statistical difference in ages in the treatment and control group.3. Name of the statistic: Standard DeviationWas this statistic covered during the class? YesNumber of times used in article: 8Subject characteristics-age:Surgery durationAnesthesia durationTimes to first flatusTotal pain medication receivedTime in rocking/non-rocking chairsTime to dischargeLaps ambulated per dayWhy do you think this statistic was used? The standard deviation is a descriptive statistic and the researchers would have used it to give a range of variance from the mean. The standard deviation describes how close the data deviates from the mean (average). A standard deviation is a measure of spread around the mean and should always be reported with it. In normal distribution a great deal of information can be obtained by measuring how many standard deviations away from the mean something is. 99.7% of data falls within 3 standard deviations from the mean (in either direction).The standard deviation needed to be used in this article to give a description of the range of values around the mean (median for age) for the different areas. Subject characteristics-age: median rocking 56.2 +/- 10.1 years: non-rocking 54.8 +/- 11.4 years Surgery duration: rocking 3.61 (hours) +/- 2.35: non-rocking 3.01 (hours) +/- 2.08Anesthesia duration: rocking 4.77 (hours) +/- 2.50: non-rocking 4.03 (hours) +/- 2.13Times to first flatus: rocking 3.16 (days) +/- 0.86: non-rocking 3.88 (days) +/-0.80Total pain medication received: rocking 29.35 (mg) +/- 58.99: non-rocking 36.48 (mg) +/- 51.66.Time in rocking/non-rocking chairs: Listed daily for 5 daysTime to discharge: rocking 7.69 days +/-4.57 days: non-rocking 7.89 days +/- 3.20 daysLaps ambulated per day: Listed daily for 5 daysWhat did this statistic show or prove: The application of a standard deviation to the categories of data collected, describe the range that the data is spread out from the mean (average, center). The standard deviation normalizes the data.Was there an associated P value? A P value was associated with the mean/SD dataIf yes, what did the P value show?Subject characteristics-age P=.600, gender-P=.218, ethnicity P=.875, marital status P=.757, and diagnosis P=.752. All not significantSurgery duration-P=.280-Not SignificantAnesthesia duration-P=.204-Not SignificantTimes to first flatus-P=.001-Highly significant for rocking group passing first flatus 16.8 hours earlier than the non-rockingTotal pain medication received-rocking P<.001 and non-rocking P<.001. This was nonnormality for both groups-A Levene’s test for homogeneity was not violated and the variances were assumed to be equal.Time in rocking/non-rocking chairs-P values ranged from .039-.650 (days 1-5). Not significant.Time to discharge-P=.837-Not significantLaps ambulated per day-P values ranged from .498-.909 (days 1-5). Not significant.4. Name of the statistic: t-testWas this statistic covered during the class? Yes For this course students did not need to calculate t-tests, but needed to understand how t-tests relate to a P value.I did find conflicting information on when to use a t-test as opposed to a z. The best explanation I found was on page 291, Table 6-1, in the Triola textbook.Number of times used in article: t-tests were used 9 timesSurgical characteristicsSubject characteristics-age, gender, ethnicity, marital status, and diagnosisSurgery durationAnesthesia durationTimes to first flatusTotal pain medication receivedTime in rocking/non-rocking chairsTime to dischargeLaps ambulated per dayWhy do you think this statistic was used? T-tests are used to compare treatment and control groups for significant differences. The treatment and control group need to be as identical as possible to support the results. T-tests give credibility to the results. A resulting P value of P=>0.05 is statistically not significant and therefore would further strengthen the belief that the treatment and control group results were based on treatment or lack of treatment, and not caused by chance.The t-test was used for this research instead of a z-score because the participants were a sample of the population and the standard deviation was a standard deviation of set of samples. In addition the samples being compared are n> 30 in each group: rocking group n=34; non-rocking group n=32.What did this statistic show or prove: t-tests relate to P valuesSurgical characteristics-ranged from P=.333-.668. Not SignificantSubject characteristics-age P=.600, gender-P=.218, ethnicity P=.875, marital status P=.757, and diagnosis P=.752. All not significantSurgery duration-P=.280-Not SignificantAnesthesia duration-P=.204-Not SignificantTimes to first flatus-P=.001-Highly significant for rocker group passing first flatus 16.8 hours earlier than the non-rockersTotal pain medication received-rockers P<.001 and non-rockers P<.001. This was nonnormality for both groups-A Levene’s test for homogeneity was not violated and the variances were assumed to be equal.Time in rocking/non-rocking chairs-P values ranged from .039-.650 (days 1-5). Not significant.Time to discharge-P=.837-Not significantLaps ambulated per day-P values ranged from .498-.909 (days 1-5). Not significant.Was there an associated P value? See aboveIf yes, what did the P value show? See above5. Name of the statistic: Confidence Intervals (CI)Was this statistic covered during the class? YesNumber of times used in article: 2Why do you think this statistic was used? The confidence interval is an inferential statistic. It is used as a point estimate correction, as each sample will not have the same point estimate. The confidence interval gives an estimated range of values for an unknown population parameter. It refers to the success of the process used to find the unknown population parameter. A confidence interval of 95% means that 95% of the time the range of values from the calculation will contain the value of the unknown population parameter.The Confidence interval was used in this article to give strength to the rocking group’s significantly shorter time from surgery to passing first flatus. The rocking were able to pass first flatus 0.7 days (16.8 hours) earlier than the non-rocking, within a 95% probability. It was also used with the results of time in a rocking chair versus time in non-rocking chairs. Those results showed no significant differences in the groups at a 95% confidence level.What did this statistic show or prove: The confidence interval shows that if this study were repeated multiple times with the same sample size, 95% (or whatever value is set for the confidence interval) of the time, the true value will be within the range.Was there an associated P value? There was not a P Value specific to the confidence interval. The P value was associated with the corresponding data and is covered in the t-test section.If yes, what did the P value show? NA6. Name of the statistic: Levene’s TestWas this statistic covered during the class? NoNumber of times used in article: 1Why do you think this statistic was used? A Levene’s test is used to test the assumption that the population variances are equal (homogeneity of variance). It checks the results of the two-sample t-tests that were used. The data results for pain medication received in both groups indicated a need for further validation. The Levene’s test is that further validation.What did this statistic show or prove: The results indicated the Levene’s test was not violated and variances were assumed to be equal.Was there an associated P value? YesIf yes, what did the P value show? The P value, P=.624 indicated no significant difference in pain medication use between the rocking and non-rocking.7. Name of the statistic: PercentagesWas this statistic covered during the class? YesNumber of times used in article: 56-27 times to the rocking group demographics and surgical characteristics, 27 times to the non-rocking group demographics and surgical characteristics and 2 times for confidence intervals.Why do you think this statistic was used? A percentage is a proportion. Proportions are a comparison of a part to the whole. It is a descriptive statistic and provides a good way to mentally visualize the data. What did this statistic show or prove: In the research article, percentages were used to break down the distribution of characteristics to the whole sample population. It was used with patient demographics to show how many subjects in either the treatment group or control group had a certain characteristic. There were 34 subjects in the treatment (rocking) group and 32 subjects in the control (non-rocking) group. The P values for the characteristics indicated no significant differences; the percentages demonstrate this on a different level.Percentages are used in research studies because they are proportions of the whole and are easy to use in comparison and expressing the results of the comparison.Was there an associated P value? The P values associated with the characteristic are associated with the percentage as well.If yes, what did the P value show? P values are explained in the standard deviation and t-test sections.8. Name of the statistic: PowerWas this statistic covered during the class? YesNumber of times used in article: 1Why do you think this statistic was used? The power is calculated prior to the start of a study by the researcher. The power of a hypothesis test is the probability of rejecting a false null hypothesis. The higher the power of a study, the less likely a type II error will be committed. A type II error rejects a true null hypothesis. A common requirement of a hypothesis test is a power of at least .80, which is the power set for this study. The null hypothesis would state there is no difference in the treatment (rocking) and control (non-rocking) groups. The power of the test is the 80% probability that there will be a significant difference between the rocking and the non-rocking in a reduction in the time from surgery to first flatus, a reduction in the total pain medication use and a reduction in time to discharge.What did this statistic show or prove: The study showed that there was a highly significant reduction in time from surgery to passage of first flatus (P=.001). There was a reduction in total pain medication use however not enough to have statistical significance (P=.604). There was not a significant difference in reduction in time to discharge (P=.837). Was there an associated P value? The power is applicable to the whole study. While there are not P values directly associated to power, I believe the P values associated with the different research areas are relevant.If yes, what did the P value show? P values are explained in the standard deviation and t-test sections.ReferencesMassey, R.L., (2010) A randomized trial of rocking-chair motion on the effect of postoperative Ileus duration in patients with cancer recovering from abdominal surgery. AppliedNursing Research, (23), 59-64. doi: 10.1016/j.apnr.2008.06.001Triola, M. M., (2006). Biostatistics for the biological and health sciences. Boston, MA. Pearson Education.Wellesley College. (2000). Independent Samples T Test. Retrieved from ................
................

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

Google Online Preview   Download