3.1. Data Collection
Strauss and Corbin Grounded theory approach was applied for a more enhanced identification of chronic pain management process in the elderly in Ahvaz during 2012-13. Grounded theory approach includes several steps that their careful execution will emerge one hidden theory on information as the outcome (
22).
Data collection was started with purposeful sampling followed by theoretical sampling. During theoretical sampling, selection of each participant was based on the gathered data from previous sample or samples (
22,
23). The selection criteria for elderly people participated in this study were as follows: ≥ 60 years of old, having experience of non-cancerous chronic pain, fully consciousness, willing to explain their emotions and experiences relevant to study subject, could speak in Persian, having efficient mental stability for transformation of experience, no history of cognitive disorders (which determined by gaining 6 score or higher in Iranian Version of the Abbreviated Mental Test Score) (
24), and having not psychological disorders, blindness, and deafness confirmed by their physicians.
In addition, elderly participants were selected with maximum variation such as age, sex, educational backgrounds, as well as socioeconomic and marital status, living with wife, widows and staying with families or relatives. The selection criteria for other participants were having a good deal of information related to chronic pain management among elderly relatives and health care providers, also willing to explain this information.
Hence, after gathering data from initial interviews with elderly people, primary categories emerged, which guided researchers for other interviews with some of their relatives and health care providers until selected persons could help for better clarifying of theory evolution.
The participants of this study were as follows: 30 elderly people over 60 years old, 3 elderly relatives, and 29 health care providers whose work were more related to chronic pain management in the elderly. The group of health care providers composed of 3 general physicians with MPH degree in geriatrics, 3 geriatric nurses, 3 psychologists, 3 physiotherapists, 3 specialist in clinical nutrition, 2 orthopedists, 2 neurologists, 1 neurosurgeon, 2 anesthesiologist, 2 specialist in clinical pharmacology, 2 psychiatrist, 1 occupational therapist, and 2 social workers.
Unstructured interviews, memo writing and observation of participants’ nonverbal behaviors were the main methods of data collection, which continued until data saturation occurred (
21). Before each interview, all participants were explained about the purpose of study and confidentiality of the information and recordings interviews. Then, the written consent was obtained with regard to willingness to participate in this study. The interviews were conducted face to face by the researcher either in the nursing homes, participant’s homes, hospitals, medical clinics or parks within Ahvaz based on the respondents’ preferences.
The interviews were conducted based on an open question and followed by probing questions to satisfy the study goals. The elderly interviews were started with open questions such as “what can you say about your pain?” and for their relatives, “what can you say about the pain of your elderly relative?” and for the health care provider, “what can you say about chronic pain management in the elderly?” Then the pursuit questions were discussed based on of participants' information that clarified the investigated concept. The next interview questions were designed on the basis of extracted categories. In addition, probes such as “Could you tell me more about that?” and “What do you mean by that?” were used to obtain more in-depth responses.
Regarding the time of each interview, Filed and Morse recommended that it should not be longer than one hour, but experience shows interview’s duration depends on interviewee (
25). In the present study, the interview with each participant was conducted in one session, which lasted 30 to 50 minutes, based on participants tolerance and interests and was recorded by a voice recorder.
In order to obtain trust and confidence in transmission of participants ' speech, interviews were transcribed verbatim exactly to their colloquial language, then typed digitally and rechecked for accuracy. Next, the texts were read several times to obtain a full perception of their concepts and immediately organized and analyzed by qualitative analyzing software MAXQDA 10. Furthermore, participants’ nonverbal behaviors during the interview were investigated, recorded and analyzed at the end of each interview. All of interviews and observations conducted by one researcher.
Content analysis of the interviews was performed according to Strauss and Corbin’s method of constant comparison (
23). All of data obtained from transcripts, observations of participants’ nonverbal behaviors and memos analyzed concurrently. Analysis of the interview data was a guide for selection of subsequent samples and sampling continued until data saturation.
Open, axial, and selective coding were used for data analysis. In the open coding process, concepts related to study were identified and coded based on two coding techniques either the in vivo codes were respondent speech or observed their status during interview or implied by inferred codes by the researcher.
In axial coding, coded data were firstly compared with each other; then, primary codes were reduced to subcategories, and the categories were developed. The similar categories were combined and compared to others with respect to their differences until more abstract categories appeared.
Finally in axial coding, the main categories with subcategories were related to each other, based on the paradigm of “causal conditions,” “context,” “controlling strategies,” and “consequence of strategies”. In the selective coding, researcher identified the variables and main concepts and selected the topics for the concepts (
21,
23).
In order to assure the data rigor, the four scales of trustworthiness recommended by Lincoln and Gubba, which consist of credibility, conformability, dependability and transferability (
26). Thus, for gaining confidence of data credibility, the researcher had longtime relation with study places, which aided him to attract participants' trust and assist to reach a better understanding of study environment.
In addition, sampling guidelines, including diversity in selecting participants and gathering data were used; this technique led to obtain more data validity based on conducted variety, extension of participants, and demographic indexes.
To determine the conformability of the results and confirm data accuracy and codes, we used revision by the participants (Member check), namely after coding process, returning the text of the interview to them for accuracy assurance of codes and explanation were conducted. Then, correction of codes would be conducted if earned codes hadn’t matched with explanatory participant’s codes.
To determine data dependability, some of the interview texts were revised by colleagues (Expert check), and as the more codes and categories emerged the research was revised by three other faculty members. This revision showed 86%-90% agreement in derived results. We used method of Polit and Hungler to compute agreement rate (
26). For instance, if the number of emerged codes in one interview by a researcher was 92 codes, and the second person agreed upon 81 codes out of the total 92 codes, this was calculated to a rate of 88.04%.
The results were discussed with samples who had not participated in this study for confirming data transferability and their viewpoints in the line of our assessed results and found as that confirmed too.
Moreover, for increasing data validity and acceptability, different methods were used
such as constant observation and investigation, designation of efficient time for data gathering, well communication with participants and conducted interviews in suitable places selected by them (such as the nursing home, participant’s home, hospitals, medical clinics or parks, which are located in Ahvaz City).