Critique of quantitative research methods and applications

¶ … Nursing

Nurse Attitude and Perception on Children's Pain

The purpose of this study made by Griffin, Polit, & Byrne (2008) was to illustrate pediatric nurses' projected responses to children's pain as described in vignettes of hospitalized children and to find out the characteristics of nurses that may possibly have an effect on those responses.

Griffin and company (2008) mentioned that the perception of the nurses to the pain that the children are experiencing do not always correspond to children's self-reports, that are regarded to be as the most important indicators of pain.

The literature review evidently says that the study was based on a study made by Rush & Harr (2001), where they did a survey of 260 experienced pediatric nurses and found less than one third of the nurses used the pain assessment tool they had developed in the institution and preprinted on the nursing flow sheet to facilitate quality assessment and pain management. In addition, nurses did not use developmentally appropriate pain assessment tools.

Griffin and colleagues (2008) have used previous research made by other agencies and organizations on pain management in a clinical setting -- research by the U.S. Agency for the Health Care and Policy Research, American Academy of Pediatrics, among others; prior studies that implied that nurses do not efficiently deal with children's pain, hence, institutions created practical norms that can give improved guarantees that the pain management of the children will be dealt with in an effective manner.

Hypothesis or the Research Questions that were Raised in the Study

They (Griffin et al., 2008) have hypothesized that nurses with higher levels of education, more clinical experience, and personal experience with intense pain would be more likely than other nurses to ascribe high levels of pain in hospitalized children and to propose giving them more pain treatments.

Because of the apparent lack of change in practice, it is important to for them to conduct research in the practice setting to attempt to uncover reasons why the use of children's pain research findings is so limited. This study focused on identification of nurse perceived barriers to providing adequate pain management in the clinical setting and asked, "What are the factors that influence children's nurses' pain assessment and management practices?" And "Were the characteristics of the nurses related to the differences in the perceived pain and the proposed pain treatments?"

Sample

Griffin and colleagues (2008) sent questionnaires to a sample of registered nurses who lives all over the United States and also to get, through the help of American Nurses Association (ANA), the names and contact information of pediatric nurses; but ANA was not capable to give a national list of registered nurses with a pediatric specialty. Consequently, Griffin and company (2008) used another source of enlisting a geographically varied sample. The population for this study was made up of registered nurses who subscribed to pediatric journals or who purchased pediatric books. The accessible population was 9,000 pediatric nurses whose names and addresses were available through the Marketing Services of Boston Company. The Griffin and company (2008) employed researchers and made them aware that these researchers were recruiting respondents that were expected to be experienced, well-educated, and well-read in their field.

Research design

Griffin and company's (2008) questionnaires were made up of three pictures or illustrations and questions about pain management for every represented child. The pictures showed an image of the child's medical condition and treatment, key physical characteristics of the child (i.e., age, weight in kilograms), what prescriptions the child received, the child's rating of pain, and the analgesic dosage prescribed by the physician should be administered as needed (PRN). In all cases, Griffin and partners (2009) said that the physicians' orders gave nurses some discretion in managing pain; nurses could theoretically administer up to 0.1 mg/kg of morphine sulphate every 3 to 4 hours. Griffin and associates (2009) also consulted a Pediatrician with specialization in pain management to guarantee the clinical validity of the orders.

Internal / external validity

The photos used were pre-tested by six pediatric nurses that gave criticisms in order to improve the photos and questions, and to fix any misconstruction in the presentation of the questionnaire. Griffin and company (2008) mentioned that the America Pain Society, in 2004, said that PRN range orders for opioid analgesics are a common clinical practice and are deemed as an appropriate and flexible method for constructing essential modifications from an individual's reaction to medication, although it was also seen as a general basis of insufficient pain management.

Data collection

The significant result that Griffin and company (2008) got from their research was using a practically large national sample - is the dominance of responses that signify awareness of the necessary medications choices connected to children's pain, in spite of nurses' characteristics. The nurses in their research formulates judgments that were equal with every child's report of pain and said that they will be able to execute sufficient pain medication by means of near to the maximum treatment recommended in addition of non-pharmacologic approaches. On the contrary of previous literature, nurse characteristics did not forecast nurses' evaluations of children's pain and need for treatment. Generally, the nurse assessed the children's pain at the same high levels at which the children reported it.

Instruments, reliability & validity

The questionnaire, that Griffin and company (2008) used, was to be able to generate the outcome variables for this research, which indicate the abstract choices of the nurses. The nurses graded their perception of the child's pain level on a 100 mm visual analogue scale with two anchor points - "No Pain at All" at one end and "Pain as Bad as Could Be" at the other. They scores were measured to the nearest millimeter, with higher values indicating greater perceived pain (Griffin et al., 2008).

Analysis for data

Griffin and partners (2008) used T-test and analysis of variance to evaluate the correlation relating nurses' characteristics and their pain management recommendations. These methods allowed the hired researchers to shoe vivid data that assisted them with their meta-analysis. On the other hand, simultaneous multiple regression was applied so they can analyze the degree that the nurses' characteristics altogether were projections of their perceived pain as well as their choices for pain management.

Every outcome measures were averaged across the three photos to reach even approximations of the nurses' attitudes toward the children's pain management; and the characteristics of the nurses in these evaluation consists of those hypotheses that might possibly be utilized from previous research such as: educational background, duration of clinical practice, and previous practice in pain management, and those that had inadequate or no earlier inquiry that is connected to child pain outcomes (nurse practitioner status, race, current nursing status, and recent receipt of pain treatment training); that few tests were statistically significant (Griffin et al., 2008). The educational background of the nurses, was found that there was no relation to the perceived pain and pain management decisions of the nurses -- for nurses without a bachelor's degree had average pain ratings, analgesic dosages, and number of non-pharmacologic methods comparable to nurses with more education, contrary to the researchers' hypothesis; and previous practice of pain, present pain training, and duration of clinical practice also were not substantially connected to some of the pain outcome measures (Griffin et al., 2008).

Altogether, variability in the nurses' pain perceptions and selected pain treatments was only weakly and inconsistently related to the characteristics that prior research suggested might play a role in pain decisions. Furthermore, Griffin and company (2008) used a multiple-comparison correction (administered when several dependent or independent are being performed simultaneously) is applied to the analyses shown in the nurses characteristics in relation to the pain perceptions and pain treatment, (to adjust for an inflated risk of a type I error when running 21 separate tests), a p value below 0.002 might be required; and no more than one link was important at this level- pain ratings and current employment status. Since they have predicted that the nurse characteristics is related with pain management outcomes, multiple regression analyses in the analysis plan were incorporated to assist in separating recognized relationships between nurses' characteristics (years of clinical experience is correlated with age). Griffin and partners (2009) said that the nurse characteristics were not beneficial variables in predicting of some pain outcomes, and the multiple regression analyses generated very low R2s; and the corrected R2 in this multiple regression analysis, even if statistically noteworthy, was only 0.04. Among other characteristics held constant, two characteristics were significantly linked to total dose - non-pharmacologic status and prior experience with pain; and those nurses who had previous personal practice of pain mentioned that they will give higher dosage of medication than those who do not have previous experience in such pain experience (means of 0.172 mg/kg versus 0.165 mg/kg, respectively) when other characteristics were restricted (Griffin et al., 2008).