Critique and analysis of quantitative research methods

¶ … Ventilator associated pneumonia (VAP) accounts for the majority of nosocomial pneumonia which may lead to more extensive hospital stay and increased intensive care. Endrotachael tubes that provide continuous subglotic suctioning (abbreviated: CSS-ETT) may reduce VAP, but they are more expensive than the standard endrotacheal tubes (abbreviated: S-ETT) that do not have the characteristic of continuous suctioning. The objective of this study (Speronni et al., 2011), therefore, was to measure the comparative costs of CSS-ETT against S-ETT among intubated people and see whether indeed the more costly CSS_ETT do show a difference that makes their expense worthwhile.

The issue that was discussed were the comparative merits of CSS-ETT compared with S-ETT and to assess whether the merits of one are more significant than the merits of another and significant to the point that their added cost is worth the hospital's investment in the resource.

2/2: Correct critique of sampling design. Cite guideline references used for critiquing sampling designs. The most reliable scientific studies follow a random sampling strategy. The study in this caase was not randomly sampled. It was conducted at a 155-bed community hospital with a combined 11-bed medical and surgical ICU. Chart review was used to collect data on 154 intubated patients that were hospitalized over a 23-month period between 2005 and 2007.

The researchers selected two groups of patients (n= 77 in each group). Patients were divided between those who received CSS-ETT and those who simply received S-ETT. All patients were 18 years and older. Exclusionary conditions included patients who were ventilated by a tracheostomy or intubated for less than 48 hours.

The researchers did note later that lack of random sampling did place certain limitations on their study. They explained, however, that lack of time, cost and other constraints impeded them from carrying out the requirements of strict scientific research, namely random sampling.

The purpose of random sampling is to prevent as much as possible subjectivity of the researcher/s and, therefore, through drawing of lots or through some other blind manner individuals are haphazardly chosen to be represented in the study.

However, there are occasions when omission of random sampling is warranted and these occasions can even make the study more scientific than had random sampling been included.

Organizers, for instance, may spend a great deal of unnecessary time and money trying to randomly attract a certain core group of people, when more qualified individuals would in reality be found were a convenience sample used. These people could then be divided into experimental and control groups and, the study being well-matched to them, would have proceeded in a more effortless manner showing possibly better results (Tuckett et al., 2001).

Even were random sampling to be as scientific as alleged, it may not always be the best situation and sometimes, the study loses by attempting to force itself into the framework of random sampling. A more advantageous approach may occur as in this case where the study employs sampling to some extent by randomly distributing its convenience sample between two groups.

Researchers elaborated that they reviewed data from charts and that patients were intubated according to the American Heart Association practice guidelines for Advance Cardiac Life Support and the hospital's standard operating procedure (SOP).

The researchers provide thorough details about the condition of each of the two groups of 77 patients. This is important so that we can replicate experiment would we so wish. They also provide lengthy details of the data and outcome that they noted as well as the humongous list of medicines that the patients were administered . It seems as though each and every detail was included. In order to take extra precaution that no Type 1 error was committed, researchers also included special statistics to screen out such a possibility.

2/2 Critique ethical considerations taken by the researcher

All researchers state regarding this is that: "This research was approved by an institutional review board. The research was conducted in accordance with the ethical standards set forth in the Helsinki Declaration of 1975" (p.16)

1/1: Rated the evidence from a hierarchy of evidence. Justify the rating of the evidence.

According to both the pyramid way of assessing evidence (Dicenso et al., 2009)and Stillwell et al., (2010) this study receives a 9 and ranks ontop of the pyramid. I only fall short from granting it a 9.5 since it is not completely randomized. It is however a control study where the experiment closley follows and represents the goal of the study and conducts its investigation in a way that as scrupulously as possible omits all possible error

2/2: Demonstrated if study was or was not valid and reliable. For reliability, addressed errors of measurement, reliability of measuring instruments, and internal consistency. Identified the type of validation used for the instruments in the study (i.e. face, content, criterion-related). If aspects of reliability or validity were not addressed in the article, this must be stated.

Study seems valid and reliable since there is internal consistency; documentation was objectivley collected from medical records and Diagnosis of VAP was based on the CDC diagnostic criteria (CDC, 2003) -- a reliable tool. The type of validation used for the instruments was both content and criterion-related. All detials were as consistent as possible for both groups furthering validity and reliability of study. Criteria, too, of each and every element employed / condiition expected were closely defined and the patient was scrupulously monitored and all data recorded in an objective, professional punctillious manner / 2: Critique statistical methods used. See chapters 16-18 in Polit and Beck (2012).

T-tests and Mann-Whitney tests were used to distinguish possible statistical differences between the two groups. T-test is the common test used; Mann is used in order to discover whether one of the two samples had larger differences than the other. (Fay & Proschen, 2010). The test was ideal for this situation. In order to discover whether categorical differences could be found, the researchers also conducted a Chi-Square and Fisher tests. Finally, a Bonferonni was conducted to screen out the possibility of the more severe Type 1 error from occuring. This may have been possible in this case due to multiple comparisons. 0.05 was placed as alpha level. 1/1: Identify computer hardware/software and discuss current version with reference citation.

SAS 9.1.3 was used. It is described as being "the leader in business analytics software and services, and the largest independent vendor in the business intelligence market." (SAS.com). Although this is a biased response, SAS is widely used as a purview of current research would indicate.

2/2: Identified other important findings and recommendations (or stated there were none); describe application to nursing.

The S-ETT group had one case of VAP; the CSS-ETT group had none. The mean total hospital charges were higher for the S-ETT group ($103,600; CSS-ETT= $88,500) (p=0.3). What we learn form this is that Although the average number of intubation days and ICU days were greater for the CSS-ETT group, there were no cases of VAP compared to the S-ETT group. Based upon the one S-ETT VAP