| dc.description.abstract | Background: Life expectancy has increased, causing an increase in the population of frail elderly. Evidence shows that elderly suffer unnecessarily because of widespread underassessment and undertreatment of their health-related problems. In palliative care patients unrelieved pain is a common problem. Effective pain management in elderly includes extra challenges such as age-normative beliefs, underreporting of pain on the part of the patients, proper assessment and atypical manifestations of pain(such as other distressing symptoms). A variety of distressing symptoms correlate with pain in palliative care patients: anxiety, fatigue, loss of appetite, insomnia and dyspnoea. Insight into these symptoms as predictors for pain may help to gain early identification of pain in independently living frail elderly in palliative care.
Aim: To determine whether the symptoms anxiety, fatigue, loss of appetite, insomnia, and dyspnoea are predictors for pain in independently living frail elderly in palliative care and to develop a prediction model.
Method: Cross-sectional study. Community-care nurses from multiple organisations included eligible patients. Utrecht Symptom Diary assessed symptom burden and Case Report Form assessed relevant covariables(age, sex and living situation).
Results: Eighty-three patients were included. Multivariable logistic regression showed dyspnoea as a predicting symptom(p=0.030) and sex(female) as a predicting covariable(p=0.047). Area Under the Curve(=0.723, p=0.001) indicated the accuracy of the final model as fair, with sensitivity of 68.1% and specificity of 66.7%.
Conclusion: This model helps with earlier identification of presence of pain through signalling of presence of dyspnoea and female-sex.
Recommendations: Early identification of pain can help community-care nurses in early advanced care planning by discussing and providing adequate (non)pharmacological pain management for independently living frail elderly. Unnecessary suffering from pain may be prevented through early identification with the use of the prediction model. | |
