The aim of the thesis was to translate, psychometrically test, and further develop the Behavioral Pain Scale for pain assessment in intensive care and to analyze if any other variables (besides the behavioral domains) could affect the pain assessments. Furthermore, the aim was to explore the patients’ experience of pain within the intensive care.
The Behavioral Pain Scale (BPS), consisting of the domains “facial expression,” “upper limbs,” and “compliance with ventilator/vocalization,” was translated andculturally adapted into Swedish and psychometrically tested in a sample of 20 patients(study I). The instrument was then further developed within one of the domains and tested for inter-rater reliability, discriminant validity, and criterion validity (study II). The method for analysis in both study I and II was a method specifically developed for paired, ordered, and categorical data. To describe and analyze the process of pain assessment, a General Linear Mixed Model was used to investigate what variables, besides the behaviors, could be associated with the observers’ own assessment of the patients’ pain (study III). Further, the patients’ experiences of pain when being cared for in intensive care were explored (study IV) through interviews with 16 participants post intensive care. Qualitative thematic analysis with an inductive approach was used for the analysis.
The first psychometric tests of the BPS (study I) showed inter-rater reliability with agreement of 85%. For the discriminant validity, all domains, except “compliance with ventilator,” indicated discriminant validity.
Therefore, in study II, a developed domain of “breathing pattern” was tested alongside the original version. The BPS showed discriminant validity for both the original and the developed version and an inter-rater reliability with agreement of 76-80%. Wheninspecting the respective domains there was a difference in discriminant validity between the original domain of “compliance with ventilation” and the developed domain of “breathing pattern,” showing higher values on the scale for the developed domain during turning. For criterion validity, the BPS showed a higher sensitivity than the observers, who on the contrary had a higher specificity.The General Linear Mix Model (study III) showed that heart rate could be associated with the observers’ assessments of pain. For the behavioral signs, the result indicated that breathing pattern was most associated with the observers’ pain assessment, whilst facial expression did not show any impact on the observers’ assessments.
The patients’ experiences of pain (study IV) in intensive care were described as generating a need for control; they experienced a lack of control when pain was present and continuously struggled to regain control. The experience of pain was not only related to the physical sensation but also to psychological and social aspects, along with the balance in the care given, which was important to the participants.
In conclusion, the translated and developed version of the Swedish BPS showed promising psychometric results in assessing pain in the adult intensive care patients. Still, other signs, besides behavioral, is possibly used when pain assessing and therefore information about and training in pain assessment are needed to enhance the assessments that are made. Also, the patients’ own experiences highlight the importance of individualizing and adapting pain assessment and treatment to the needs of each patient. Making them a part of the team could enhance their feeling of control, thereby supporting them in facing the experience of pain.