Eria, and confirmed simultaneously by three people: the patient’s self-report; the assistant doctor; and a relevant family member. The assistant doctor was blinded, in other words, not involved with data collection, data analysis or writing the manuscript. Individuals in the sample were selected with the aid of their medical doctors specialized in the treatment of eating disorders. Potential informants were told about the study by their doctors. They were provided with written information about the study. All participants gave their written consent. The second strategy involved “snowball”, in which selected participants could suggest other `information rich’ subjects [30]. Patients with acute psychotic symptoms, mental impairment, cognitive deficits or a certain speech or auditory impairment that could compromise communication with the researcher were not included. In addition, individuals with alcohol or drug abuse or dependence were not included if acutely intoxicated. Socioeconomic data were obtained through [31], allowing classification into five classes, from A (individuals with the highest income level: above 15 minimum wages/month [US 4,030.00]) to E (those with the lowest income level: equal to or below K minimum wage/month, [US 134.00]).coding scheme. Frequent group discussions helped PHCCC increase agreement with the coding system and led to the development of a thematic structure, including both manifest (explicit) and latent themes. Coding was compared and differences of opinion resolved through examination of the text. Cohen’s Kappa interrater reliability ranged from .70 to 1.0 for each thematic item. All constructs were validated against the original text using confirmatory and selective coding and following the `top-down principle’ [33]. Data was collected after approval by the Ethics Committee (1468/08), of the Federal University of Sao Paulo.ResultsFifteen women with remitted AN were interviewed between November 2008 and May 2009. Most of the participants were young 22948146 women, with early onset AN (15 to 24 years old), were mostly single or divorced, with a high education and socioeconomic level. One participant achieved remission for ten years. See table 1. All participants (n = 15) had undergone treatment with psychotropic medication (selective serotonin reuptake inhibitors); some had psychotropics plus psychotherapy (n = 12); some had combined treatment with a nutritionist (n = 10); some had alternative treatments combined with drugs or psychotherapy (n = 4). Three patients needed hospitalization. Search for treatment occurred between 6 to 18 months of the index episode. Participant’s descriptions of their remission experience revealed several preliminary categories: personal factors; external factors; treatment factors. Each component has multiple dimensions. See table 2. By condensing the preliminary categories that contain a description of an experience that the informants identified as contributing to the remission process we were able to identify four major high order constructs. As we are willing to collect the participants own experiences associated with remission, these four higher order constructs are built on a bottom-up structure, in other words, this is a set of information brewed by the patients themselves. The following are the four constructs: 1) motivation and stimuli for remission; 2) buy Sermorelin empowerment/autonomy; 3) media related factors; and 4) treatment factors. See table 3.Interview TechniqueTo elicit the women.Eria, and confirmed simultaneously by three people: the patient’s self-report; the assistant doctor; and a relevant family member. The assistant doctor was blinded, in other words, not involved with data collection, data analysis or writing the manuscript. Individuals in the sample were selected with the aid of their medical doctors specialized in the treatment of eating disorders. Potential informants were told about the study by their doctors. They were provided with written information about the study. All participants gave their written consent. The second strategy involved “snowball”, in which selected participants could suggest other `information rich’ subjects [30]. Patients with acute psychotic symptoms, mental impairment, cognitive deficits or a certain speech or auditory impairment that could compromise communication with the researcher were not included. In addition, individuals with alcohol or drug abuse or dependence were not included if acutely intoxicated. Socioeconomic data were obtained through [31], allowing classification into five classes, from A (individuals with the highest income level: above 15 minimum wages/month [US 4,030.00]) to E (those with the lowest income level: equal to or below K minimum wage/month, [US 134.00]).coding scheme. Frequent group discussions helped increase agreement with the coding system and led to the development of a thematic structure, including both manifest (explicit) and latent themes. Coding was compared and differences of opinion resolved through examination of the text. Cohen’s Kappa interrater reliability ranged from .70 to 1.0 for each thematic item. All constructs were validated against the original text using confirmatory and selective coding and following the `top-down principle’ [33]. Data was collected after approval by the Ethics Committee (1468/08), of the Federal University of Sao Paulo.ResultsFifteen women with remitted AN were interviewed between November 2008 and May 2009. Most of the participants were young 22948146 women, with early onset AN (15 to 24 years old), were mostly single or divorced, with a high education and socioeconomic level. One participant achieved remission for ten years. See table 1. All participants (n = 15) had undergone treatment with psychotropic medication (selective serotonin reuptake inhibitors); some had psychotropics plus psychotherapy (n = 12); some had combined treatment with a nutritionist (n = 10); some had alternative treatments combined with drugs or psychotherapy (n = 4). Three patients needed hospitalization. Search for treatment occurred between 6 to 18 months of the index episode. Participant’s descriptions of their remission experience revealed several preliminary categories: personal factors; external factors; treatment factors. Each component has multiple dimensions. See table 2. By condensing the preliminary categories that contain a description of an experience that the informants identified as contributing to the remission process we were able to identify four major high order constructs. As we are willing to collect the participants own experiences associated with remission, these four higher order constructs are built on a bottom-up structure, in other words, this is a set of information brewed by the patients themselves. The following are the four constructs: 1) motivation and stimuli for remission; 2) empowerment/autonomy; 3) media related factors; and 4) treatment factors. See table 3.Interview TechniqueTo elicit the women.