Authors' Response

The authors thank Dr Roe for her commentary on this article, particularly in highlighting opportunities for further research in this vulnerable and understudied population of long-term catheter users. She suggested continuing research in this area related to: (1) perceptions and decision making related to symptoms of catheter associated infections; (2) objective measures of UTI vs subjective symptoms; (3) stratified samples related to catheter type (suprapubic vs urethral), length of use, purpose of catheter, and medical diagnoses; and (4) other catheter-related issues, such as the use of catheter valves (not approved for use in the United States by the Food and Drug Administration). Dr Roe summarized some of the major findings and directed readers to consider implications of the findings related to the design. We would like to respond to several points in her remarks about the research design to clarify how we conducted this study.

First, the authors view the research as a descriptive and exploratory study that was quantitative in its methodological approach, research traditions, and philosophic underpinnings. Although narrative data were used from the tape-recorded interviews, these data were viewed as supplemental sources to verify accuracy between what was heard and what was written during the interview. Portions of the tape-recordings were preselected to identify information related to the major catheter concerns, such as episodes of catheter-related problems, such as UTI, blockage, or difficulty in insertion or removal of the catheter. Narrative data also were used to provide more detailed description about sensations and practices related to living with a urinary catheter. Narrative data and numeric data were analyzed iteratively to deepen knowledge about possible relationships among the major variables, such as fluid intake, urinary output, catheter blockage, and UTI. This information was categorized and organized in contingency tables. Miles and Huberman1 and patterns were explored for expected relationships based on research literature. Finally, narrative data were analyzed descriptively to organize practical tips suggested by participants for catheter management.

Content analysis was originally used for consumer research. The purpose was to essentially count and categorize responses based on a predetermined set of questions or parameters. Over time, content analysis became a useful method for narrative data in either quantitative or qualitative studies. Content analysis for our study began with an a priori approach of predetermined and well-defined parameters. Qualitative research, by its nature, is far more open-ended in its approach, and its purpose often includes searching for meanings, new theories, and a richness and depth of description that is not possible in quantitative studies.

We did not see this study as an exploration of patients' experiences of living and coping with an indwelling catheter, as stated by Dr Roe in her commentary. The PI previously conducted a qualitative study that explored this experience using phenomenologic research methods and the philosophic approach of Merleau-Ponty, related to embodiment (what we know through our bodies). The qualitative design in that study opened new avenues for research and provided in-depth glimpses into the experiences of catheter users, both of which are consistent with the purpose of qualitative studies. In contrast, the narrative data in the current study was used to further understand issues about urine flow in catheter users that were uncovered in the qualitative study-fluid intake, urinary output, UTI, and catheter blockage-and subsequently explored through a chart audit. The researcher in the current study used well-defined parameters and multiple approaches to understanding relationships among these variables.

Accordingly, some of the points Dr Roe raised must be viewed with caution. For instance, generalizability is related to how well the sample population represents the larger population of interest. The central limit theorem suggests that a sample size of 30, which the current study had, is usually adequate for statistical tests and the generalization related to those tests if the sample is normally distributed for the given variables.2 Dr Roe stated that "qualitative research uses small carefully chosen samples because of the type and amount of data analysis" and that "qualitative data is more difficult to generalize to other catheterized groups." Those are accurate statements, but this study was not a qualitative study; furthermore, with a sample of 30, the study was large enough for the generalizability associated with the Chi square testing. The authors would agree with Dr Roe that information from the narrative data should be acknowledged as case-related data and not intended for inferences to the larger population of catheter users.

Finally, the authors would like to thank Dr Roe for raising some important issues about further research and about measurement of UTI and studies in community settings. UTI research in this population is exceedingly limited. Most research has been done in hospitalized persons with short-term catheters. Long-term catheter use has been studied in populations with SCIs, but the focus has varied. Most often, studies are designed to determine long-term effects on the urinary system, such as pyelonephritis or bladder cancer. People with a long-term catheter inserted for more than 30 days will consistently have bacteriuria. Moreover, bacteriuria is almost never treated with antibiotics, but rather only symptomatic UTI is treated. Therefore, the authors disagree with the suggestion to study perception of bacteriuria in comparison with symptomatic UTI. We do not believe that people are likely to perceive bacteriuria once colonized; however, it may be helpful to know what people perceive in the early and later stages of an episode of symptomatic UTI and how they go about making the decision to seek medical care.

References