Readmission of patients with diabetes is a problem that warrants consideration of the contributing factors. Readmission of patients within 30 days of discharge is considered to be an indicator of healthcare quality -- along with other circumstances, such as patient lifestyle -- that needs to be addressed from a patient care perspective and from a cost of care perspective (Dungan, 2012). A dismal statistic starkly represents the problem: Roughly 8% of the U.S. population is represented by patients with diabetes, yet this group accounts for 23% of the hospitalizations in the nation (Dungan, 2012). On top of this figure, between 14.4% to 21% of diabetic patients are readmitted, compared to 8.5% and 13.5% of U.S. hospital patients overall (Dungan, 2012). The problem is exacerbated by the rise in national rates of diabetes means that more patients will present from the general population and, accordingly, more patients with diabetes will experience readmission -- unless the cycle of admission-readmission is prevented and absolute numbers are reduced. Rubin, et al. (2014) explored the causes of readmission that occurred within 30 days of discharge among diabetic patients, a phenomenon the researchers refer to as early readmission. The purpose of the qualitative research was to understand the conditions that led to readmission from the perspective of the patients with diabetes, with an emphasis on "aspects of the inpatient to outpatient transition of care" (Rubin, et al., 2014, p. 1). Rubin, et al. (2014) developed their research to answer questions about the reasons for early readmission through the use of qualitative research. The research question is broadly stated and lends itself well to qualitative research, as the study is grounded in the qualitative tradition of personal accounts (interviews) and the philosophical orientation of epistemology: the study of knowledge. The study was designed to circumvent the constraints of previous research that depended upon retrospective, lack of clinical evidence-based data, and were restricted to "primary discharge diagnosis of diabetes," which was confirmed through a review of the hospital records (Rubin, et al., 2014, p. 2).

The subjects were included in the study if they met the following inclusion criteria: eligible English-speaking patients with diabetes, who had been readmitted to the hospital within 30 days of discharge, and were at least 18 years of age. The patients were selected from Temple University Hospital, which is an urban academic medical center located in Philadelphia, Pennsylvania. The criteria for a diagnosis of diabetes included: ICD-9-CM code of250.xx, self-report, or outpatient use of a diabetes-specific medication. The list of potential participants, which excluded obstetrics patients, was generated through the use of an electronic medical record system. Of the original sampling frame, 6 patients declined to participate and 1 record was lost, resulting in a sample of 20 participants interviewed between September 2012 and February 2013. These procedures are consistent with qualitative research methodology.

A convenience sample was then identified and approached for consent, using the human subjects review process required by the Temple University Institutional Review Board. It is important to note that the study outcomes are intended to provide benefit to the participants, which is one of the tenets of ethical conduct provided in the Belmont Report (Polit & Beck, 2012). The Belmont Report argued for three standards of ethical conduct in research: Beneficence, respect for human dignity, and justice (Polit & Beck, 2012). Each of these principles is important, but the beneficence holds the most relevance for the study under review, such that the researchers are compelled to minimize harm and maximize benefit to the research participants or to others (Polit & Beck, 2012). The ethical treatment of study participants includes the provision of adequate information to establish informed consent, which means that participants understand they can refuse to participate or terminate their participation at any time without giving a reason. Informed consent also requires that participants understand how the data will be used, how long it will be retained, what happens to the data once the study ends or is published, what form it will take when shared, and that aggregate data will be preferred and individual data will not be associated with any personal identifier related to the study participants.

Data was collected from participants who had given informed consent through 20-minute interviews that were conducted in their hospital rooms. Three investigators independently conducted interviews, transcribed the interviews, and triangulated the interview narratives with discharge records to ensure the patients' recall accuracy. Given that the study is perhaps one of the first to use qualitative methods to explore readmission of patients with diabetes, the investigators developed an interview guide derived in part...

The 28-question interview protocol contained open-ended items in order to capture the actual words participants used when responding to questions asked by the interviewers. A pilot was conducted with three participants, but no changes were made to the interval protocol or guide following the pilot, so the pilot interviews were included in the study data. The interview process provided for data saturation, as interviews were conducted until thematic saturation was accomplished and no new information was likely to be derived from additional interviews. Moreover, the interviewers probed for more information when the study participants answered the questions with just "yes" or "no" responses. The interview protocol covered a multitude of topics, including ability to attend follow-up appointments, discharge instructions, home environments, needs after discharge, social support, transitions of care between the inpatient and outpatient settings, and other general questions about their experiences and perceptions of the current and recent hospitalizations.
Rigor was ensured in the data analysis process as each investigator coded their interviews individually, and then the group of investigators collectively reviewed each of the interviews in order to assure the coding was accurate, representative, and consistent. The data analysis process was entirely appropriate for the data collected since qualitative data is generally coded for emerging themes with related elements of the data being categorized as they are encountered in the review (Dye, et al., 2000). This process is sometimes referred to as the constant comparison method (Dye, et al., 2000). The process of coding the interviews according to emerging themes is appropriate to and consistent with qualitative data analysis techniques (Dye, et al., 2000). The thematic analysis indicated five themes that reflect contributions to the risk of readmission. The themes were as follows: 1) Poor health literacy, health system failure; 3) failure of expected protective factors; 4) social determinants of health impeding care; and 5) loss of control over illness (Rubin, et al., 2014, p. 4). Overall, the study participants had limited or mistaken information about their disease or about the factors that influenced the progression of their diabetes. Many participants had poor recall about their discharge and were unclear about important variables such as changes in medication or if follow-up medical visits were scheduled. Discharge instructions were found to lack information specific to diabetes even though all the participants had diabetes. Participants described situations that they perceived as interfering with their ability to follow discharge instructions. These conclusions are supported by the research conducted by Hellman (2014), which found that readmission rates could be reduced by "improved transitions of care, including medicine reconciliation, detailed education of the patient at the time of discharge, and appropriate, patient centered education and instruction for both the patient and the medical team caring for the patient" (p. 4).

A lack of support following discharge resulted in participants being readmitted, and many of these issues were fundamental, such as an inability to obtain insulin, medication, meals, or nursing care. Donnell-Jackson, et al. (2014) found similar results in a study about contributing factors to readmission of patients with diabetes, concluding that "the DC [discharge] process does not successfully communicate instructions to readmitted patients and instructions often fail to address diabetes" (p. 1). In Rubin, et al. (2014), diabetes management was insufficient to accomplish good glycemic control and many participants had an HbA1c>8%. While social support was considered by participants to be adequate, it typically was not and the same obstacles seemed to reoccur. Indeed, participants were often unable to obtain their meds, get transportation to medical appointments, or prepare their own meals. Many participants were discouraged by their illness and felt inadequate to the challenge of managing their life situations and their diabetes. This attitude often resulted in non-adherence to medicine, blood testing, and diet regimens. The contributing factors represented in the thematic breakdown were observed to interact in such a way that the presence of more than one variable could result in more extensive difficulties and challenges that the participants were unable to overcome. One of the most important findings was the degree to which feelings of loss of control over chronic illness appeared to contribute to readmission risk.

The investigators identified several areas of future research that are entirely relevant since the current research is recent. Future research is needed to confirm the themes identified and to determine if other diseases result in similar risk for readmission. The recommendations made by the researchers also need to…