Support Care Cancer (2012) 20:1667–1671 DOI 10.1007/s00520-011-1257-y
ORIGINAL ARTICLE
Development of a knowledge assessment tool for dermatotoxicity caused by inhibitors of epidermal growth factor receptor Dimitrios Konstantinos Papageorgiou & Theocharis Konsatntinidis & Nektaria Skandalaki & Anna Papadouri & Theodora Pappa & Stavroula Poulopoulou
Received: 24 November 2010 / Accepted: 24 August 2011 / Published online: 18 September 2011 # Springer-Verlag 2011
Abstract Purpose The aim of this study was the development of a knowledge assessment tool for dermatotoxicity caused by inhibitors of EGFR (intravenous regimens). Methods Five nurses with experience in oncology created a 25-item questionnaire. The questionnaire was presented to six experts for assessment of face and content validity. Item
D. K. Papageorgiou (*) Euroclinic Athens, Athens, Greece e-mail:
[email protected]
analysis and reliability testing were evaluated on the test results of 76 nurses. Results Face and content validity was achieved for 25 items. Two items with low biserial correlations were deleted. The values for item difficulty range from 0.2 to 0.7. The values for item discrimination ranged from 0.25 to 0.64. The complete post-tested 23-item questionnaire showed excellent internal consistency with Kuder–Richardson 20 score of 0.909. The Cohen κ tests showed that the questionnaire has very good test–retest reliability. The specific tool can be used in several studies, leading to the development of educational interventions.
T. Konsatntinidis Department of Nursing, Technological Institute of Heraklion, Heraklion, Greece e-mail:
[email protected]
Keywords EGFR . Dermatotoxicity . Oncology nurses . Assessment tool
N. Skandalaki Venizeleio Hospital of Heraklion, Heraklion, Greece e-mail:
[email protected]
Introduction
A. Papadouri : T. Pappa Oncology Hospital Ag. Anargiri, Athens, Greece A. Papadouri e-mail:
[email protected] T. Pappa e-mail:
[email protected] S. Poulopoulou Department of Statistics, Athens University of Economics and Business, Athens, Greece e-mail:
[email protected]
In recent years, the use of monoclonal antibodies has entered the active battle against cancer, either as monotherapy or in combination with other chemotherapeutic agents. The epidermal growth factor (EGFR) belongs to a family of four closely related cell surface receptors: EGFR (HER1 or erbB1), HER2/ neu (erbB2), erb3 (HER3), and erb4 (HER4). After the receptor has been activated, the signal through an intracellular transport system reaches the cell core, leading to the activation of genes and promotion of cell differentiation [1]. EGFR (intravenous regimens) is expressed or overexpressed in many different solid tumors such as head and neck cancers, non-small cell lung cancer, colon cancer, pancreatic cancer, ovarian cancer, cervical cancer, and stomach cancer [2, 3]. The EGFR inhibitors that have been
1668
approved are cetuximab (Erbitux) and panitumumab (Vectibix). Both substances have specific ways of preparation and administration, a similar mode of action, and similar side effects [4, 5]. The majority of patients treated with EGFR inhibitors (intravenous regimens) are likely to experience skin side effects [6, 7]. Patients suffer from exfoliative dermatitis and acute generalized pustular dermatitis resembling common acne [5, 8]. The management of this side effect requires initial assessment and then specific therapeutic interventions [7]. Simultaneously, we must also consider the effect on the quality of life as it is a toxicity that causes a change in body image [9]. Aim of the study The aim of the present study was the development of a knowledge assessment tool for dermatotoxicity caused by EGFR inhibitors (cetuximab, panitumumab). Significance of the study Nurses as equal members of the therapeutic team are actively involved not only in the preparation, storage, and administration of the regimens but also in patient support. Knowledge directly affects the provided health care, both qualitatively and quantitatively. All international nursing organizations proposed as more accurate and more appropriate the evidence-based nursing practice. Unfortunately, new therapeutic agents are developed rapidly and require constant updating and knowledge. Through this research, we will try to develop an assessment tool for dermatotoxicity caused by the EGFR inhibitors (intravenous regimens). The use of the specific tool will lead to results that will help us design future educational and clinical interventions in order to improve nurses’ knowledge in the clinical oncology environment. This is the first study done in Greece.
Support Care Cancer (2012) 20:1667–1671
questions). Answers for the 25 assessment questions were correct, incorrect, and don't know. Content validation The questionnaire was presented to six experts for assessment of face and content validity. All experts had at least a master's degree in nursing or medical social sciences and had at least 3 years of experience in an oncology care unit. To achieve face validity, the experts were asked whether all questions were clearly worded and would not be misinterpreted. From the evaluations by the panel of experts, the content of the questionnaire was found to be relevant, accurate, complete, and clear. For content validity, the experts evaluated the nursing relevance of the 25 items using a scale of 1 to 4, where 1=not relevant, 2=somewhat relevant, 3=quite relevant, and 4=highly relevant. Per item, an content validity index (CVI) [10] that reflects the proportion of the consulted experts agreeing on the content validity of an item was calculated. When six or more experts are consulted, the CVI must be at least 0.83, reflecting 1 as disagreement. Description of the sample and item analysis Participants and sample characteristics A convenience sample was recruited from eight urban healthcare institutions with an oncology department in Greece over a 3-month time frame. The intention was to include all nurses who met the eligibility criterion. The eligibility criterion was that participants must be registered general nurses working in oncology units; 100 nurses met the eligibility criterion, and 76 returned completed questionnaires. The majority of the nurses (92.1%) were females and graduated from the Technological Educational Institute (64%). Of the nurses, 42.1% were working at day clinic units and 38.2% at oncology departments; 94.7% had no specialization in oncology. The mean age of the sample was 38(±6%)years. The median of working years in oncology department was 4.5, with range of 28.
Methods Item analysis Method of instrument development Item development Five nurses with experience in oncology created a questionnaire which was separated into three parts. The first part included the demographic data (age, education level, working years, and specialization). The second part included questions about frequency in using the inhibitors and the patient's training. The third part included the assessment questions (25
The answers in the test are true, false, and don't know. Every answer in the full 25-item scale was coded “1” if the respondent provided the correct answer and “0” if an incorrect or “don't know” answer was given. Individual scores were expressed as the sum of the correct answers. Biserial correlations (i.e., a form of the Pearson productmoment statistic used when one variable is dichotomous) of each separate item with the total score of the scale were computed to determine unidemensionality between the
Support Care Cancer (2012) 20:1667–1671
1669
Table 1 Item analysis Question
No. of answers (%)
Table 1 (continued) Item Item Corrected item, difficulty discrimination total correlation
Q1 True
0.33 a
33 (43.4)
Question
Q13
0.4
True Falsea 0.30
27 (35.5) Truea False 22 (28.9) Don't know 27 (35.5)
0.6
Q4 True
0.39
30 (39.5) 28 (36.8) Falsea Don't know 18 (23.7)
Q5 Truea
43 (56.6)
False
9 (11,8)
Don't know 28 Q7 True 35 14 Falsea Don't know 27
0.31
0.41
0.3
0.2 0.53
0.33
0.57
0.7 0.65
0.7
0.38 38 (50) 9 (11.8)
0.35
0.4
0.25 52 (68.4) 1 (1.3) 23 (30.3)
0.43
0.3
0.44 8 (10.5) 51 (67.1) 17 (22.4)
0.33
0.5
0.58
0.62
0.46
0.52
0.40
0.71
0.54
0.65
0.64
0.41
0.48
0.62
0.43
0.65
0.25
0.61
0.31
0.38
0.35
0.70
0.35
0.65
0.5
0.5
0.6
0.6
0.5
Don't know 21 (27.6) Q18 Truea False
0.46 (46.1) (18.4) (35.5)
49 (64.5) 6 (7.9)
Don't know 21 (27.6) Q17 39 (51.3) Truea False 16 (21.1)
(36.8)
Q8 25 (32.9) Truea False 12 (15.8) Don't know 39 (51.3) Q9 28 (36.8) Truea False 16 (21.1) Don't know 32 (42.1) Q10 True Falsea Don't know Q11 Truea False Don't know Q12 Truea False
0.61
0.31
False 4 (5.3) Don't know 23 (30.3) Q16 Truea False
0.6
19 (25) 39 (51.3)
Falsea Don't know 18 (23.7) Q15 49 (64.5) Truea
0.33
0.4 0.35
Don't know 24 (31.6) Q6 24 (31.6) Truea False 24 (31.6)
0.62
16 (21.1) 39 (51.3)
Don't know 21 (27.6) Q14 True
0.38
48 (63.2) Truea False 22 (28.9) Don't know 4 (5.3)
0.26
0.4
Q3
Item Item Corrected item, difficulty discrimination total correlation
Don't know 29 (38.2)
0.34
False 22 (28.9) Don't know 21 (27.6) Q2
No. of answers (%)
39 (51.3) 12 (15.8)
Don't know 25 (32.9) Q19 38 (50) Truea False 20 (26.3) Don't know Q20 True Falsea Don't know Q21 True Falsea Don't know Q22 Truea False
0.5
18 (23.7) 4 (5.3) 48 (63.2) 24 (31.6) 31 (40.8) 25 (32.9) 20 (26.3) 52 (68.4) 8 (10.5)
Don't know 16 (21.1) Q23 True 7 (9.2) 48 (63.2) Falsea Don't know 21 (27.6) a
0.5
Correct answer
0.6
0.3
0.7
0.6
1670
Support Care Cancer (2012) 20:1667–1671
items if they are homogeneous and unidimensional. Items with low biserial correlations (r < 0.25) were deleted. Revising tests on the basis of their scores is an essential part of improving instruction [11]. Therefore, the difficulty level and the discriminative value of the items of the questionnaire were analyzed. The difficulty level of a question is defined as the proportion of respondents who answered the question correctly [12]. Possible values range from 0.0 to 1.0. Items that are answered correctly by more than 90% of the respondents (value
