Patient and Provider Perspectives on Postsurgical Recovery of Cushing Syndrome

Rachel Acree; Caitlin M. Miller; Brent S. Abel; Nicola M. Neary; Karen Campbell; Lynnette K. Nieman


J Endo Soc. 2021;5(8) 

In This Article


This study had 2 phases. The first was an open-ended patient survey designed to identify themes in patients' responses regarding challenges during postsurgical recovery and helpful coping mechanisms. The second was broader in scope and targeted both patients and physicians; themes identified in the first phase were given as possible responses to questions about challenges and coping mechanisms, and open-ended questions also were included. Because no personally identifiable information was collected, the Intramural National Institutes of Health Office of Human Subjects Research Protections determined the study to be exempt from institution review board evaluation.

Phase 1 Survey

In the first phase, patient members of the Cushing's Support and Research Foundation (CSRF) (approximately 400 members), were invited to complete an open-ended survey about their experiences after surgical treatment of CS in late 2008. The 7-question survey asked about the type of CRM, the duration of its use and tapering/testing strategies for its discontinuation, and the experience of recovery including both challenges and helpful coping mechanisms, as shown in Table 1.

Ninety-three responses were collected. Two respondents were excluded for answering only the first question, leaving 91 patients included in the analysis. All patients reported having resection of a specific lesion followed by initiation of CRM; this was considered confirmation of CS remission. For this report, 1 author (R.A.) reviewed all open-ended responses and clustered them into thematic groups.

Phase 2 surveys

In the second phase, we developed 2 web-based surveys based on responses of the phase 1 open-ended survey. One survey targeted patients in remission after surgical treatment of CS and was promoted through CSRF newsletter and emails to approximately 915 patients from December 2010 to September 2011. Responses were collected from December 2010 to November 2011. This 27-question survey asked about the cause of CS; type and time of surgery; the type; dose and duration of CRM; decisions about discontinuation of CRM; postsurgical symptoms; whether and how patients received information about the experience during recovery; time to return to work; what helped during recovery; and length of recovery (Table 2). Answers were selected using a drop-down menu; additional open-ended responses could be added. A final open-ended question asked for any additional feedback and what might have improved the recovery experience.

A total of 385 patients submitted responses. Forty-four submissions were excluded from analysis: 2 were cases of exogenous CS, 2 patients had not yet had surgery, 1 was a duplicate, 1 was in the immediate postoperative period, 3 were taking ketoconazole, 1 was awaiting repeat surgery, 7 answered ≤ 3 questions, and 27 included only demographic information. As a result, 341 patients were included in the analysis, of which 332 reported having resection of a specific lesion followed by initiation of CRM. Of the 9 who did not indicate that they received postoperative cortisol replacement, 3 stated that cortisol levels were normal, 1 had adrenal insufficiency after surgery, 1 had undergone unilateral adrenalectomy, and 5 listed no medications or additional treatments. We considered these 341 patients to be in remission after surgery.

A second survey targeted endocrinologists. It was promoted through mailings to adult endocrinology fellowship training programs in the United States and the United Kingdom, and through an email announcement by the Pituitary Society. Responses were collected from December 2010 to November 2011. This 22-question survey asked about the type of practice and number of CS patients seen annually and to date; the type; dose and duration of CRM; decisions about discontinuation or taper of CRM based on minimal dose and testing; the type and delivery of information about the recovery period; the time to discontinuation of CRM and to full recovery; patients' overall recovery experience; and what suggestions they make to improve that experience (Table 3).

Sixty-one endocrinologists responded to the physician survey. Seven responses included only demographic information and were excluded, leaving 54 responses included in the analysis.

Survey responses were obtained using an anonymized, secure, online platform (Clinical Trials DataBase, hosted by the National Institute of Child Health and Human Development), accessed via a link provided in the outreach notifications.

Statistical Analysis

No respondent answered every survey question, so the total number of responses for each question varied in both the patient and physician surveys. In general, only questions with response rates > 80% were analyzed. Exceptions to this included if the question had a lower response rate because it was not applicable to every respondent or if it provided important contextual information. Such cases are noted. Open-ended responses were reviewed and clustered into thematic groups; when themes in open-ended responses directly answered closed-ended questions in the survey that were left blank by the respondent, responses were aggregated. When this methodology was used, it is noted when presenting the results.

Descriptive statistics included mean or median results with SD or interquartile range (IQR). One outlier response about time to recovery (158 months) was excluded because it seemed unlikely to be correct. The percentage of responses was calculated based only on questions that were answered; because not all respondents answered all questions, the number (n) of responses differed. Open-ended text answers were organized by themes and counted. Unpaired t tests, χ2 analysis, Mann-Whitney U tests, Wilcoxon signed-rank tests, or 2-proportion z tests were used for comparisons where appropriate. P values and 95% CIs are reported. P values were adjusted using the Bonferroni correction in all χ2 post hoc analyses.