Symptoms of older orthopedic and rheumatic patients

The study has been approved by the appropriate ethics committees. Informed consent was obtained from all patients included in the study. Additionally, the respective responsible physicians gave their written consent to study participation.

We performed a cross-sectional study by using a semi-structured telephone interview.

The participants were recruited in a university hospital’s three orthopedic wards (treatment focus on endoprosthetics, reconstructive joint interventions, and general traumatology) and one rheumatological ward (treatment focus on rheumatic diseases). Older patients admitted to one of these wards were invited to participate during their hospital stay. Inclusion criteria were age at least 70 years old (to enrol older patients), living independently, taking at least 5 drugs (to enrol patients with polymedication to treat multimorbidity), sufficient cognition assessed by the hospital physician, and written informed consent. Additionally, the patients needed to be able to talk on the telephone (adequate hearing and language skills).

We used a quantitative explorative semi-structured telephone interview to assess the extent and variety of symptoms including those that might not have been communicated before. This design should facilitate the participation of older multimorbid patients in a telephone survey.

Two pharmacists with experience in palliative and geriatric pharmacy developed a semi-structured interview based on Somers et al. [22], the Memorial Symptom Assessment Scale [17] and the Common Terminology for Adverse Events [15]. Physicians with experience in geriatrics were involved for consultation.

To ensure comprehensibility and feasibility, the interview guide was cognitively pretested (by paraphrasing and comprehension probing) stepwise with nine professionals (pharmacists not included in the study design) and with three laypersons (older patients within the inclusion criteria). The participants in the pretest were not involved in developing the study protocol or the main study. The results of the pretests were not included in the final data assessment. The pretest with the professionals resulted in improved comprehensibility and notes to adjust the interview duration. The lay pretest resulted in the following modifications: inclusion of the symptom “hearing impaired”, shortening the interview guide (not asking which activities of daily living were impaired by the symptom, asking about treatment for only the 5 most severe or impairing symptoms), simplifying the wording in questions to increase lay understanding, including the response categories “ward physician” and “nobody” for the question about who recommended treatment.

The final interview (Supplement 1) consisted of four parts: (i) symptom characteristics (occurrence, intensity, impairment, timing, and cause), (ii) symptom communication, (iii) symptom treatment, and (iv) further support.

During the first week after discharge, the interviewer (pharmacists) contacted the patients by telephone to make an appointment for the telephone interview. The interviewer wrote patients’ answers on a documentation sheet and documented the interview’s duration. For quality assurance, the interview was audio-recorded. Complete data were digitalized within 24 h of the interview. To ensure complete, accurate and comprehensible data, a second study person randomly checked every sixth data set for transcription errors and inconsistency. No irregularities were found. Data were analyzed descriptively. Open answers were categorized post hoc.

A total of 60 patients (35 women and 25 men) completed the telephone interview (Fig. 1). We included 50 patients from the orthopedic wards and 10 from the rheumatological ward. In the median, the patients were 77 years old (min-max: 70–91 years) (Table 1). We performed the interview in median 12 days after discharge (Q25/Q75: 10/15 days; min–max: 5–37 days), which took 38 min (Q25/Q75: 31/50 min;min-max: 14–143 min).

After their discharge from the hospital, the patients reported up to a maximum of 14 symptoms. Most of them already existed before hospitalization. Besides the expected symptoms of restricted mobility and pain, these patients were also affected by various other symptoms, such as insomnia or dry mouth. The fact that approximately one third of the symptoms were of severe intensity and impairment indicates a need for more effective treatment strategies, including deprescribing. For 16% of the reported symptoms, the patients considered drug treatment as the main cause, which might indicate adverse drug reactions. This is below the 28% rate reported in other studies [19]. Polypharmacy increases the risk of an adverse drug reaction, but nonspecific symptoms hamper its recognition [13]. Therefore, it can be assumed that the patients have considered only some of the possible existing drug causes for their symptoms.

Although one third of the symptoms were not communicated by the patients before, they still desired improvement. “Wait and see” was a frequent answer regarding symptom treatment in our study, which is in line with the fact that older patients tend to wait or ignore their symptoms or manage them by themselves [1, 3]. Whitaker et al. provided examples for future interventions in symptom appraisal in cancer patients, which might be transferable to older multimorbid patients. They distinguished symptom knowledge, attention, expectation and identity [24]. Further measures are urgently needed to empower patients in self-management [12]. Measures should include the patients’ social and physical environment [5] as well as instrumental and emotional support [9].

One fifth of the most alarming symptoms were not treated, which could indicate a gap in symptom management. Patients must first become aware of their physical changes and then discuss their symptoms with a healthcare professional [20]. The patients in our study seemed to lack sufficient motivation to communicate some of their symptoms. As the patients were interviewed after hospital discharge, however, they already had recent contact with physicians. They also considered it relatively easy to obtain support and actively ask for it. Therefore, it has to be questioned how these consultations can be followed by better overall symptom diagnosis and treatment. This treatment includes deprescribing, especially in the case of adverse events. It can be supposed that many symptoms were not addressed during inpatient treatment because of the complexity of older multimorbid patients. Ekdahl et al. generated a theory that a lack of time, competence and a holistic view hamper good care for older hospital patients [8]. For our respondents, pharmacies were the easiest to reach healthcare institutions but were less actively asked for support. This could be because the public is reluctant to trust pharmacy services [10]. Nevertheless, pharmacies counsel on self-medication to treat minor ailments and are interested in intensified interprofessional collaboration with other healthcare professionals [11]. It should be investigated how this low-threshold contact point can additionally support referring older patients to healthcare structures.

The number of 60 participants only allows limited conclusions about occasionally mentioned symptoms (limited generalizability). The nature of the interview as a telephone call resulted in the exclusion of patients with severe hearing problems. Our survey captured only the patients’ perspectives, which might differ from a professional assessment. The interview was quantitative; therefore, in-depth analysis could not be done. The interview guide was pretested but not validated. To reduce reporting and confirmation bias, we presented answers to all questions in the interview guide. Even though patients were asked to add further symptoms, it has to be considered that the number of 24 predefined symptoms influenced the number of reported symptoms. We included typical geriatric symptoms and frequent adverse events and only asked the patients questions regarding symptom treatment for the 5 most severe and impairing symptoms. We did not further differentiate between grades of clinical relevance. Although we consider deprescribing to be an equal treatment option, the interview did not specifically address this issue. Response bias, social desirability, and recall bias cannot be excluded.