Prevalence of Factors Contributing to Unplanned Hospital Readmission of Older Medical Patients

The objective of this cross-sectional survey study to identify unique factors and aspects involved in the process of unplanned hospital readmissions of medical patients aged 65 or above within 30 days of discharge.

To identify these factors following research questions will be examined:

1. what are the most prevalent factors that contribute to 30-days readmissions among medical patients above 65 years?
2. to what degree do patients, relative and health care professionals agree upon the contributing factors?
3. where on the patient pathway between sectors are the contributing factors most prevalent?

Unplanned Readmission is defined as follows:

• The admission occurs between four hours and 30 days after the last admission
• The onset of the admission is acute
• The admission cannot be a result of a referral from another hospital or ward
• The patient cannot have a cancer- or accident diagnosis
• Admission not part of existing treatment regimen, such as intravenous antibiotic treatment, or readmission planned at latest discharge.

Respondents are divided in five groups: patients and their relative, community-based nurses, general practitioners (GP) and hospital-based doctors.

DATA COLLECTION Patients will be prospectively recruited September 2020 to June 2021. All eligible patients will be asked to sign an informed consent for use of individual healthcare data. No formal ethical approval is necessary. Data collection will be performed by the PhD student and project staff.

Data will be collected from two sources; 1) from questionnaires and from 2) CROSS-TRACKS.

Data from questionnaires There is no existing questionnaire that examine factors that contribute to readmissions among medical patients aged 65 or above. Therefore, to evaluate unique dimensions on the same readmission, study-specific questionnaires have been developed to examine the perspectives on the readmission from the patient, relative, community-based nurse (if the patient has home health care), GP, and receiving hospital-based doctor. The development process will be described below.

Development of the questionnaires:

In short, the stepwise process of developing the study-specific questionnaires is described below.

Step 1: to prepare and conduct semi-structured interviews to generate knowledge from the target groups. In total, 25 interviews of one hour were conducted with patients, relatives, GPs, community-based nurses and hospital doctors. These interviews were transcribed and the content was condensed and coded using NVIVO.

Step 2: to gain knowledge from peer reviewed evidence on risk factors and predictors of readmission among older medical patients.

Step 3: the 3 questionnaires were developed. Knowledge gathered from the two preliminary steps were used to create the individual items in the questionnaire. A formative questionnaire model was chosen. Likert scales, open-ended questions, multiple choice and yes/no/don't know were chosen as response formats, developed, and subsequently validated by an expert in questionnaires. The questionnaires were pilot tested twice.

Step 4: the 3 questionnaires were finalized, implemented and pilot tested in a study-specific REDCap database.

The questionnaires to GPs, hospital doctors and community-based nurses are identical. The 3 questionnaires cover following factors that can contribute to unplanned hospital readmissions: 1) disease specific factors, 2) diagnostic specific factors, treatment and care specific factors, 3) social network specific factors, 4) organisation specific factors, 5) communication specific factors, 6) competences and knowledge specific factors, 7) resources specific factors and 8) practical specific factors. The questionnaire to the patients is a short version modified to frail and vulnerable older people. The questionnaire to the relatives is an intermediate version between the questionnaire to the health care professionals and patients. All responders receive a personal link which ensures that the survey only can be answered by this specific person.

Unplanned readmitted patients will be approached and enrolled in the study within 72 hours after readmission.

Patients who accept to participate in the study, accept data collection through registers, questionnaire and accept data collection from their relatives, GP, community-based nurse and hospital-based doctor.

Data from CORSS-TRACKS Data will be retrieved from CORSS-TRACKS. CORSS-TRACKS data are gathered from different Danish sources and cover; 1) Primary care; Municipality-based electronic care provision record, Danish National Health Service Register, Danish National Database of Reimbursed Prescriptions, Prehospital electronic record, 2) Secondary care; Danish National Patient Registry, Electronic health record, Clinical laboratory information system, 3) Danish Civil Registration System and 4) Danish Register for Evaluation of Marginalization.

CORSS-TRACKS data is merged with data collected from questionnaires to conduct sub-analysis.

ANALYSIS Quantitative analysis

• Descriptive analysis including prevalence
• Cohen's kappa or similar analysis for the level of agreement between responses

Qualitative analysis

• Quantitative content analysis will be used as method to analyze open-ended questions. NVivo will be used when coding the content.