Minimal clinically important difference in days at home up to 30 days after surgery
Summary
Patient-centred outcomes are increasingly recognised as crucial measures of healthcare quality. Days alive and at home up to 30 days after surgery (DAH30) is a validated and readily obtainable patient-centred outcome measure that integrates much of the peri-operative patient journey. However, the minimal difference in DAH30 that is clinically important to patients is unknown. We designed and administered a 28-item survey to evaluate the minimal clinically important difference in DAH30 among adult patients undergoing inpatient surgery. Patients were approached pre-operatively or within 2 days postoperatively. We did not study patients undergoing day surgery or nursing home residents. Patients ranked their opinions on the importance of discharge home using a Likert scale (from 1, not important at all to 6, extremely important) and the minimum number of extra days at home that would be meaningful using this scale. We recruited 104 patients; the survey was administered pre-operatively to 45 patients and postoperatively to 59 patients. The mean (SD) age was 53.5 (16.5) years, and 51 (49%) patients were male. Patients underwent a broad range of surgery of mainly intermediate (55%) to major (33%) severity. The median minimal clinically important difference for DAH30 was 3 days; this was consistent across a broad range of scenarios, including earlier discharge home, complications delaying hospital discharge and the requirement for admission to a rehabilitation unit. Discharge home earlier than anticipated and discharge home rather than to a rehabilitation facility were both rated as important (median score = 5). Empirical data on the minimal clinically important difference for DAH30 may be useful to determine sample size and to guide the non-inferiority margin for future clinical trials.
Introduction
Measurement of healthcare quality has received increased attention over the last 30 y [1, 2]. This has been driven by efforts to improve outcomes, reduce harm and increase cost effectiveness of healthcare and the need to undertake objective and meaningful research in order that this can be achieved. Early definitions of quality used outcome measures such as morbidity and mortality rates and length of hospital stay. These measures, although important, have become less useful due to the low complication rates of anaesthesia.
Outcomes that include quality of recovery and return to pre-morbid functional status may be more important measures of healthcare quality to patients after surgery. Days alive and at home up to 30 days after surgery (DAH30) has been validated as a patient-centred outcome measure [3]. It is easy to measure and captures key aspects of the patient journey, including length of stay, re-admission, discharge destination and early death. However, we have little understanding as to what constitutes the minimal clinically important difference (MCID) for DAH30. This includes the minimum number of additional days at home that patients would perceive as beneficial or, conversely, the minimum number of additional days in hospital that patients would view as detrimental after surgery [4, 5]. We conducted a survey of patients undergoing surgery to determine the MCID for DAH30.
Methods
The study was approved by the Alfred Hospital Human Research Ethics Committee. We recruited adults (age ≥18 y) undergoing inpatient surgery. We did not study patients undergoing day surgery, nursing home residents (or those expected to be discharged to a nursing home postoperatively) and patients who were unable to provide consent. All participants were provided with a written patient information sheet. We designed and administered a 28-item survey aiming to determine the MCID for DAH30 (see online Supporting Information, Appendix S1). In addition to patient characteristic data, patients were asked about their self-reported level of frailty, prior experiences of surgery, priorities for recovery after the current surgery and when they expected to be discharged home.
Patients reported the importance of timely discharge home, avoiding hospital re-admission and discharge to a rehabilitation hospital on a Likert scale from 1 (not important at all) to 6 (extremely important). Patients reported the number of additional days in hospital that would be of importance and significance to them should there be a complication or requirement for rehabilitation, and the number of additional days at home following surgery that would be meaningful to them. This was reported using an ordinal scale (half an extra day; one extra day; two extra days; three extra days; four or more extra days). Finally, patients reported their views on a range of outcomes following surgery on a Likert scale from 1 (completely unacceptable) to 6 (extremely acceptable).
The survey was completed pre-operatively in the anaesthesia pre-admission clinic or within 2 days postoperatively. We aimed to recruit over 100 participants, based on previous studies that have enrolled 40–100 subjects [6, 7], and allowing for missing data.
Patients ranked their preferences for discharge home, hospital re-admission and discharge to a rehabilitation facility using a Likert scale from 1 (not important at all) to 6 (extremely important). Patients ranked the minimum number of additional days at home that would be meaningful to their recovery using an ordinal scale (half an extra day; one extra day; two extra days; three extra days; four or more extra days). This scale was also used to determine the minimum number of additional days in hospital due to complications and the minimum number of days in rehabilitation that would be meaningful to them.
Statistical analysis was performed using SPSS version 25.0.0.1 (IBM Corporation, Armonk, NY, USA).
Results
We approached 125 patients between July and October 2018, of which 21 declined. Data from 104 patients were analysed. The median reported clinical frailty scale score was 3 with 85 (82%) patients self-reporting as very fit, well or managing well (Table 1). Nineteen (18%) patients self-reported as being either vulnerable to frailty, or being mild, moderately or severely frail. Of the 22 patients living alone, 21 had relatives or friends to assist with postoperative care if needed. Twenty-five (24%) patients were primary carers for dependents, of which 16 were children and 8 other relatives or friends.
| n = 104 | |
|---|---|
| Age; y | 53.5 (16.5) |
| Sex; male | 51 (49%) |
| Highest level of education | |
| No formal education | 1 (1%) |
| Primary school | 2 (2%) |
| High school | 32 (31%) |
| TAFE/certificate | 22 (21%) |
| University degree | 45 (43%) |
| Doctorate | 2 (2%) |
| Currently employed | 59 (57%) |
| Currently driving a car | 89 (86%) |
| Clinical frailty scale | 3 (2–3 [1–6]) |
| Living status | |
| Lives alone | 22 (21%) |
| With spouse and/or children/parents | 62 (60%) |
| With children | 5 (5%) |
| With parents | 5 (5%) |
| With other friend/relative | 10 (10%) |
| Carer for a primary dependent | 25 (24%) |
| Pet owner | 55 (53%) |
| ASA physical status | |
| 1 | 16 (15%) |
| 2 | 39 (38%) |
| 3 | 42 (40%) |
| 4 | 7 (7%) |
| Current smoker | 16 (15%) |
| Comorbidities | |
| Hypertension | 35 (34%) |
| Obesity | 25 (24%) |
| Diabetes | 22 (21%) |
| Arrhythmia | 5 (5%) |
| Chronic obstructive pulmonary disease | 11 (11%) |
| Ischaemic heart disease | 16 (15%) |
| Heart failure | 10 (10%) |
| Chronic kidney disease | 15 (14%) |
| Chronic liver disease | 5 (5%) |
| Peripheral vascular disease | 2 (2%) |
| Chronic neurological disease | 2 (2%) |
| Prior stroke or transient ischaemic attack | 8 (8%) |
- TAFE, technical and further education.
Ninety-one (88%) patients had undergone surgery in the past, of which one-third had experienced one or more postoperative complications following this previous surgery (Table 2). Twenty (19%) patients expected to be discharged the day following surgery; 30 (29%) patients 2–3 days postoperatively; 26 (26%) patients 4–5 days postoperatively; 19 (19%) patients 6–10 days postoperatively; and 7 (7%) patients >10 days postoperatively.
| n = 104 | |
|---|---|
| Extent of current surgery | |
| Minor | 13 (13%) |
| Intermediate | 57 (55%) |
| Major | 34 (33%) |
| Urgency of current surgery | |
| Elective | 76 (73%) |
| Non-elective | 28 (27%) |
| Timing of survey | |
| Pre-operative | 45 (43%) |
| Postoperative | 59 (57%) |
| n = 91 | |
|---|---|
| Extent of previous surgery | |
| Minor | 29 (28%) |
| Intermediate | 46 (44%) |
| Major | 16 (15%) |
| Timing of previous surgery | |
| Previous 12 months | 25 (27%) |
| 1–2 years ago | 15 (17%) |
| 2–5 years ago | 17 (19%) |
| 5–10 years ago | 17 (19%) |
| > 10 years ago | 17 (19%) |
Among the priorities for recovery following surgery, surviving surgery and recovering without complications ranked most highly (Table 3). Twenty-six patients did not rank surviving surgery as their most important priority, and eight ranked surviving surgery as the ‘least important’ priority.
| Surviving surgery | 1 (1–2 [1–6]) |
| Recovering without complications | 2 (2–3 [1–5]) |
| Being pain free | 3 (2–3 [1–6]) |
| Returning home as soon as possible | 4 (3–4 [1–5]) |
| Returning to work or usual activities | 5 (4–6 [1–6]) |
| Being discharged from hospital as soon as possible (even if this means spending time in rehabilitation) | 6 (5–6 [1–6]) |
- n = 20 responses were missing or excluded due to incomplete data.
Timing of discharge home within an expected timeframe was rated as ‘quite important’ (median score, 3) but discharge home earlier than anticipated was rated as ‘very important’ (median score, 5). If discharge could be achieved earlier than expected (question 23), the median number of days at home that would be meaningful to patients was 3 days. If a complication occurred requiring prolonged hospitalisation (question 17), the median number of days in hospital or rehabilitation that would be of concern to the patient was 3. The median MCID for DAH30 were similar among the pre-specified sub-groups (see online Supporting Information, Table S1).
Management of postoperative complications at home (e.g. with home nursing visits) rather than being re-admitted to hospital was rated as ‘moderately important’ (median score, 4). If a complication occurred after discharge home, avoiding hospital re-admission was rated as ‘moderately important’ (median score, 3). If a complication occurred that required re-admission to hospital, the median number of days in hospital that would be of concern to patients was 3. Discharge home rather than to a rehabilitation facility was rated as ‘important’ (median score, 5). If inpatient rehabilitation was required (question 26), the median number of days in rehabilitation that would be of concern to patients was 3.
Patient perspectives on complications following surgery were also evaluated (question 28). Complications resulting in death and long-term nursing care were ranked most unacceptable while complications resulting in eventual discharge home were consistently ranked more acceptable (Table 4). In total, 83 (84%) patients ranked complications leading to death as ‘completely unacceptable’ and 67 (68%) ranked complications leading to discharge to a nursing home as ‘completely unacceptable’. In contrast, four patients ranked complications leading to death as ‘extremely acceptable’; five ranked complications leading to discharge to a nursing home as ‘extremely acceptable’; 11 ranked complications leading to ongoing disability as ‘extremely acceptable’; and 19 ranked a complication leading to the requirement for ongoing help at home as ‘extremely acceptable’.
| A complication leading to death | 1 (1–1 [1–6]) |
| A complication leading to long-term nursing home care | 1 (1–2 [1–6]) |
| A complication requiring ongoing hospitalisation 1 month postoperatively | 2 (1–4 [1–6]) |
| A complication resulting in not being able to return to baseline level of activity (but eventually returning home) | 3 (2–4 [1–6]) |
| A complication requiring ongoing care at home | 3 (2–5 [1–6]) |
- n = 5 responses were missing or excluded due to incomplete data.
Discussion
The median MCID for DAH30 was 3 days among a broad range of patients undergoing elective or inpatient surgery. This MCID was consistent across a range of scenarios, including earlier discharge home, complications delaying hospital discharge and admission to a rehabilitation hospital. Among patients undergoing a wide range of inpatient and elective operations, DAH30 has a bimodal distribution with a median of 23.7 days [3]. The MCID is important in order to interpret the results of clinical studies and to guide study design for clinical trials, including the non-inferiority margin [8] and sample size [9]. The MCID may be determined empirically by surveying patients, or using distribution-based methods [4]. These include the standard deviation (SD)/2 rule [10], 0.3 SD and 5% range, to estimate the MCID for quality of life instruments [11]. Applying such distribution-based methods to a previous dataset [3] leads to bootstrapped MCID estimates for DAH30 of 3.2, 2.0 and 1.5, respectively.
Quality measures should be important, scientifically acceptable, easy to use and feasible to measure [12]. The DAH30 measure has construct validity [3] and is readily obtainable from most hospital datasets. Patients in our survey ranked discharge home sooner than expected, and discharge home rather than to a rehabilitation facility as important measures of their postoperative recovery.
Of note, 26 respondents did not rank surviving surgery as their most important priority, and three patients ranked complications leading to death as ‘completely acceptable’. These responses may indicate misunderstanding of the question’s ranking system. However, free-text comments from these respondents regarding other factors important to their recovery included ‘getting back to a normal life and having no disabilities’, ‘getting back to what contributes to my quality of life’ and ‘being independent’. These comments indicate that certain patients may place more value on quality rather than on quantity of life. In contrast, five patients indicated that discharge to a nursing home would be a ‘completely acceptable’ outcome, and this may reflect the broad spectrum of values that different patients hold regarding peri-operative goals of care.
Our study has several limitations. We assessed the MCID for DAH30 using an ordinal scale with a range from half an extra day to four or more extra days. Our results may have differed had patients been asked to select the minimal important number of days using a continuous scale. We did not include non-English speaking patients who could not complete the survey in English, and this may limit the generalisability of our findings to patients from different cultural backgrounds. While our study included a broad range of patients undergoing different types of surgeries, the MCID for DAH30 may vary depending on the specific patient population or hospital setting.
The DAH30 measure is an important patient-centred outcome tool to measure peri-operative quality of care. Among a broad range of patients undergoing inpatient surgery, the MCID DAH30 was 3 days.
Acknowledgements
The authors thank M. Pollock for her assistance with the survey and database design. No external funding or competing interests declared.
