A pilot project to test the feasibility of automated text messaging to collect multi‐day patient‐reported outcomes related to pain interference after total joint arthroplasty in veterans

Acute postoperative pain trajectories may differ between surgical procedures [1] and individuals who undergo the same procedure [2]. To date, clinical measurement of acute pain has been generally limited to pain intensity, which provides an inadequate assessment of recovery [3]. Despite recommendations to incorporate patient-reported outcomes into peri-operative pain assessment and management [3], technological and other barriers have prevented incorporation into routine clinical care. Using an automated short message service text messaging system described previously for peripheral nerve block follow-up [4], we attempted implementation of patient-reported pain intensity and interference measurement for one week following total knee arthroplasty (TKA) and total shoulder arthroplasty (TSA) as a clinical care improvement project and tested the feasibility of this as a data collection tool. Over 8 months, a convenience sample of patients who underwent primary TKA and TSA at a single Veterans Affairs hospital were invited to participate. All patients owned a mobile telephone and expressed willingness to respond to pain-related questions once daily for postoperative days 1–7 as part of clinical care. Retrospective review of the data was performed with IRB approval and waiver of informed consent. All patients received multimodal analgesia with oral paracetamol and celecoxib before surgery, if not contraindicated, and a perineural catheter appropriate for site which infused 0.2% ropivacaine 6 ml.h-1 postoperatively with 5 ml patient-controlled boluses available every 30 min. Patients undergoing TKA received spinal or general anaesthesia and all patients undergoing TSA received general anaesthesia. To assess pain interference with activity, sleep, mood and stress in addition to pain intensity over the previous 24 h, we utilised the Defence and Veterans Pain Rating Scale and supplemental questions [3, 5]. These Likert-type questions (0 = no pain/interference, 10 = maximum pain/interference) were programmed securely within REDCap (Nashville, TN, USA) for data collection and integrated with Twilio (San Francisco, CA, USA) for automated text messaging at the same time each day [4]. The daily text messages and code are available in online Supporting Information Appendix S1. Any response triggered the instrument, whereas no response triggered a reminder at 15.00. Given the small sample, data are presented as descriptive statistics and no comparisons were made. From December 2021 to July 2022, 16 patients participated. Median (IQR [range]) age of patients having TKA (n = 10) was 70 (68–74 [63–79]) y, and patients having TSA (n = 6) was 50 (41–62 [27–73]) y. One TSA patient was female, and all other patients were male. Duration of stay was 3.5 (3–4 [2–8]) days for patients having TKA and 1 (1–1 [1–3]) days for patients having TSA, with 1 day being same-day discharge. The proportion of text messaging respondents per postoperative day, in aggregate and divided by inpatient vs. outpatient status, is shown in Figure 1. The highest response rate from all participants was 69% (11/16) on postoperative day 3. On postoperative days 1 and 4, 62% (10/16) responded. By postoperative days 6 and 7, only 44% (7/16) and 50% (8/16) of participants responded, respectively. This project provides some important observations related to the challenges of collecting patient-reported outcomes in routine clinical practice. Despite using the common communication modality of text messaging [4], the rate of non-responders was high during the 7-day follow-up period, and some of the responses were incomplete or invalid (e.g. text response in lieu of numerical score) despite pre-operative teaching. Response rates may have been higher in outpatients in the early postoperative period. We speculate that inpatients may have had limited access to their mobile phones or viewed the text messaging questions as redundant. However, the outpatients in this pilot were predominantly younger people who underwent TSA. Our small convenience sample, low overall recruitment rate and high frequency of attempted patient contact (i.e. daily) prevent us from drawing any definitive conclusions. We also recognise that alternative technological innovations now exist for gathering patient-reported data such as smartphone applications [6, 7] and that there may be preferences for or against the use of technology that may vary by age group or patient population [8]. The results of this pilot project suggest that automated daily text messaging may not be a reliable means for collecting patient-reported outcomes in routine clinical care. As the demand for these data increases with the trend towards patient-centred care, further research is needed to identify the most acceptable frequency and mode of contact for patients while balancing the potential burden of clinician time and effort involved in measurement. A portion of this material is the result of work supported with resources and the use of facilities at the Veterans Affairs Palo Alto Health Care System (Palo Alto, CA, USA). The contents do not represent the views of the Department of Veterans Affairs or the United States Government. ERM is an Editor of Anaesthesia. No external funding or competing interests declared. Appendix S1. Daily text messages and code. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.