gms | German Medical Science

Background and state of research: In Germany, nearly half of the population dies in hospitals, mostly outside specialised palliative care units. In order to improve the quality of care in the dying phase, national and international guidelines provide quality indicators and outcome variables for optimal care. These can be used for the evaluation of interventions to improve care in the dying phase.

Research question and objectives, hypothesis: The aim was to develop a tool for data extraction from medical records to assess the quality of hospital care in the dying phase.

Method: A data extraction tool was developed by operationalising items based on quality indicators and key recommendations on care in the dying phase from the German national evidenced-based guideline Palliative care for patients with incurable cancer [1]. The tool was then applied to extract data from n=400 medical records of deceased patients of ten general wards and intensive care units of two university hospitals.

Results: The final tool consisted of 59 items related to eight domains:

1.

(Withdrawal of) medications and measures in the dying phase;

2.

Involvement of specialist palliative care teams;

3.

Progress documentation;

4.

Continuity of care;

5.

Symptom assessment;

6.

Principles and practical issues of care in the dying phase;

7.

Artificial nutrition and hydration,

8.

Dialysis, ventilation, tumour-specific therapy.

The created items proved to be applicable for all domains on medical records of deceased patients. Information density on the different domains varied considerably within medical records. On average, most documented information could be found on the domain (withdrawal of) medication and measures (94, 1% of records) followed by the domain artificial nutrition and hydration (69, 9% of records). Fewest indications could be found in the area of principles and practical issues, e.g. shared decision-making (43,7% of records), followed by the domain progress notes e.g. physicians’ documentation of patients’ dying process (36,2% of records).

Discussion: The developed data extraction tool allows analysing, if quality indicators and key recommendations on care in the dying phase are reflected within medical records. The tool can be used in general wards as well as in intensive care units, regardless of the hospital department and the primary disease of the deceased patient. As medical record analysis is limited to the information documented, it may not provide a complete picture of the actual care provided to patients.

Implication for research: The tool enables a structured retrospective analysis of the routine documentation in medical records and the extent to which recommendations and quality indicators regarding care in the dying phase are reflected therein. To gain a full understanding of the care provided to patients in this phase, it should be combined with other methods [2].

Funding: Innovationsfonds/Versorgungsforschung; 01VSF19033