Table 2.
Overview of the studies investigating the relationship between preoperative frailty and early postoperative RC-related outcomes, grouped by endpoints of interest
| Reference | Study design | Study size, n | Type of surgery | Preoperative frailty indicator(s) | Postoperative outcome(s) | Findings |
|---|---|---|---|---|---|---|
| Postoperative complication rates | ||||||
| Palumbo et al., 2019 [22] | Retrospective (NIS database) | 23,967 | ORC, LRC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | Overall complications | Frail patients vs non frail: 67.9% vs 55.8%, P < 0.001 Frail patients vs non-frail: OR 1.54 (95% CI 1.44–1.65), P < 0.001 |
| Michel et al., 2019 [23] | Retrospective (NRD) | 9459 | ORC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | ICU-level complications (CCS = IV) | Frail patients vs non-frail: 52.9% vs 18.6%, +Δ34.3% (95% CI 30.0–40.6%), P < 0.001 Frail patients vs non-frail: OR 4.74 (95% CI 3.60–6.25), P < 0.001 |
| Pearl et al., 2017 [24] | Retrospective (ACS-NSQIP Database) | 4330 | RC | 11-item mFIa | Any in-hospital complication; Major in-hospital complicationse |
mFI, any in-hospital complication: frail (mFI ≥0.27), 28.7%; pre-frail (mFI 0.09–0.18), 24.8%; robust (mFI = 0),19.6%; P < 0.001 mFI, major in-hospital complications: frail, 19.2%; pre-frail, 12.6%; robust, 8.8%; P < 0.001 |
| Meng et al., 2018 [25] | Retrospective (ACS-NSQIP Database) | 1516 | RC | 11-item mFIa | 30-day any adverse events; 30-day serious adverse events |
mFI, 30-day any adverse events: AUC 0.500, 95% CI (0.471–0.529), did not reach statistical significance mFI, 30-day serious adverse events: AUC 0.516, 95% CI (0.483–0.550), did not reach statistical significance |
| Woldu et al., 2018 [29] | Retrospective (monocentric study) | 346 | ORC, RARC | 11-item mFIb | 30-day major complications (CCS ≥III) | mFI: AUC 0.551, 95% CI (0.471–0.631), P = 0.2 |
| Chappidi et al., 2016 [26] | Retrospective (ACS-NSQIP Database) | 2679 | RC | 11-item mFIb | 30-day severe complications (CCS IV–V) | Frail and pre-frail patients (mFI ≥2) vs the others (mFI <2): 14.6% vs 8.3%, P < 0.001 Pre-frail patients (mFI = 2) vs non-frail (mFI = 0): OR 1.84, 95% CI (1.28–2.64), P = 0.001 Frail patients (mFI ≥3) vs non-frail (mFI = 0): OR 2.58, 95% CI (1.47–4.55), P = 0.001 |
| Sathianathen et al., 2018 [27] | Retrospective (ACS-NSQIP Database) | 5516 | RC | 5-item sFIc and 11-item eFIb | 30-day major complications (CCS ≥III) | Frail (sFI ≥3) and pre-frail patients (sFI = 2) vs non-frail (sFI = 0): 26.0% and 15.1% vs 7.7%, P < 0.001. Pre-frail patents vs non-frail: OR 1.73, 95% CI (1.32–2.26) Frail patients vs non-frail: OR 3.22, 95% CI (2.01–5.17) sFI: AUC 0.56, 95% CI (0.47–0.57) eFI: AUC 0.56, P = 0.7 |
| De Nunzio et al., 2019 [30] | Retrospective (multicentre study) | 117, aged ≥80 years | RC | 5-item sFIc | 90-day major complications (CCS ≥III) | Frail patients (sFI ≥3) vs other patients (sFI <3): 11.1% vs 3.4%, P = 0.02 Frail patients vs other patients: OR 3.10, 95% CI (0.70–13.70), P = 0.01 |
| Taylor et al., 2019 [28] | Retrospective (ACS-NSQIP Database) | 92,999 (3823,8154, 14,668, 2817, 13,953, 5678, 9466) | MIRP, RRP, MIRN, ORN, MIPN, OPN, RC |
5-item sFId | 30-day any complications; 30-day major complicationse |
In RC patients: sFI 30-day any complication: sFI = 0, 50.8%; sFI = 1, 54%; sFI = 2, 61.2%; and sFI ≥3, 66.4%; P < 0.001 sFI 30-day major complications: sFI = 0, 45.3%; sFI = 1, 49.0%; sFI = 2, 55.3%; and sFI ≥3 60.3%; P < 0.001 |
| Lascano et al., 2015 [21] | Retrospective (ACS-NSQIP Database) | 41,681 (5709, 7791 1443, 23,350,3388) | PN, RN, RNU, RP, ORC | 15-point mFI and 11-item mFIa | 30-day CCS = IV complications | In RC patients: Frail patients (15-point mFI ≥0.20) vs non frail (15-point mFI 0–0.05): 17.0% vs 6.6%, P < 0.001 15-point mFI: AUC 0.585, P < 0.001 The 15-point mFI was superior to the 11-item mFI in all the comparisons |
| Burg et al., 2018 [31] | Prospective (monocentric study) | 123, aged ≥65 years | ORC, RARC | FFC | 30/90-day any complications; 30/90-day major complications (CCS ≥III) |
Shrinking, any 30-day complication: OR 3.79 (95% CI 1.64–9.26), P = 0.002 Physical activity, 30-day major complications: OR 0.36 (95% CI 0.12 − 0.78), P = 0.04 Physical activity, 90-day any complication: OR 0.84 (95% CI 0.69–1.00), P = 0.03 Intermediately frail or frail vs non-frail, 30-day major complications: OR 4.87 (95% CI 1.39–22.77), P = 0.02 Intermediately frail or frail vs non-frail, 90-day major complications: OR 3.01 (95% CI 1.05–9.37), P = 0.04 |
| LOS and costs | ||||||
| Palumbo et al., 2019 [22] | Retrospective (NIS database) | 23,967 | ORC, LRC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | pLOS; THCs |
Frail patients vs non-frail, pLOS: RR 1.32 (95% CI 1.28–1.35), P< 0.001 Frail patients vs non-frail, THCs: +$8003.3 (95% CI 6849.1–9158.2), P< 0.001) |
| Michel et al., 2019 [23] | Retrospective (NRD) | 9459 | ORC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | LOS; Hospital-related costs |
Frail patients vs non-frail, LOS: median (IQR) 15 (9–21) vs 7 (6–10) days, P< 0.001 Frail patients vs non-frail, LOS: OR 0.58 (95% CI 0.50–0.66), P< 0.001 Frail patients vs non-frail, hospital-related costs: median (IQR) $39 665 ($28 196–$56 397) vs $27 307 ($21 145–$36 049), P< 0.001 Frail patients vs non-frail, hospital-related costs: OR 0.42 (95% CI 0.34–0.49), P< 0.001 |
| Pearl et al., 2017 [24] | Retrospective (ACS-NSQIP Database) | 4330 | RC | 11-item mFIa | LOS | Frail patients (mFI ≥0.27) vs robust (mFI = 0): 9 (4–57) vs 7 (0–75) days, P< 0.001 |
| Meng et al., 2018 [25] | Retrospective (ACS-NSQIP Database) | 1516 | RC | 11-item mFIa | pLOS (>75th percentile) | mFI: AUC 0.529 (95% CI 0.498–0.560), did not reach statistical significance |
| Woldu et al., 2018 [29] | Retrospective (monocentric study) | 346 | ORC, RARC | 11-item mFIb | LOS; Hospital-related costs |
Frail (sFI ≥3) and pre-frail patients (sFI = 2) vs non-frail (sFI = 0), mean (SD) LOS: 11.3 (11.1) and 8.2 (5.7) vs 7.6 (4.7) days, P = 0.003 Frail patients (mFI ≥3) vs other groups, hospital-related costs: $30 354 vs ~$22 500, P = 0.003 |
| Chappidi et al., 2016 [26] | Retrospective (ACS-NSQIP Database) | 2679 | RC | 11-item mFIb | LOS | Pre-frail and frail patients (mFI ≥2) vs other patients (mFI <2): 11.0 (10.2–11.8) vs 10.0 (9.7–10.4) days, P < 0.01 |
| De Nunzio et al., 2019 [30] | Retrospective (multicentre study) | 117, aged ≥80 years | RC | 5-item sFIc | LOS | Frail patients (sFI ≥3) vs other patients (sFI <3): 16 (10–23) vs 13 (10–20) days, P = 0.28 |
| Discharge disposition | ||||||
| Michel et al., 2019 [23] | Retrospective (NRD) | 9459 | ORC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | Non-home vs home discharge | Frail patients vs non-frail: 33.9% vs 11.6%, +Δ22.2% (95% CI 16.2–14.5%), P < 0.001 Frail patients vs non-frail: OR 3.43 (5% CI 2.50–4.69), P < 0.001 |
| Pearl et al., 2017 [24] | Retrospective (ACS-NSQIP Database) | 4330 | RC | 11-item mFIa | Discharge to non-home vs home care | Frail patients (mFI ≥0.27) vs robust (mFI = 0), if experienced a major in-hospital complication: 54% vs 22%, P < 0.001 Frail patients vs robust, if not experienced a major in-hospital complication: 23% vs 7%, P < 0.001 Pre-frail patients (mFI 0.09–0.18) vs robust: OR 1.37 (95% CI 1.07–1.74), P = 0.01 Frail patients vs robust: OR 2.33 (95% CI 1.34–4.03), P = 0.003 |
| Meng et al., 2018 [25] | Retrospective (ACS-NSQIP Database) | 1516 | RC | 11-item mFIa | Discharge to a higher level of care vs at home | mFI: AUC 0.586 (0.536–0.636), did not reach statistical significance |
| Sathianathen et al., 2018 [27] | Retrospective (ACS-NSQIP Database) | 5516 | RC | 5-item sFIc and 11-item eFIb | Discharge to a facility vs home | Frail (sFI ≥3) and pre-frail patients (sFI = 2) vs non-frail (sFI = 0): 27.8% and 18.7% vs 8.4%, P < 0.001 Pre-frail patients vs non-frail: OR 1.54 (95% CI 1.18–2.02) Frail patients vs non-frail: OR 2.31 (95% CI 1.40–3.82) sFI: AUC 0.60 (95% CI 0.58–0.62) eFI: AUC 0.58, P = 0.5 |
| Unplanned readmission | ||||||
| Michel et al., 2019 [23] | Retrospective (NRD) | 9459 | ORC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | 30-day readmissions; Readmission costs |
Frail patients vs non-frail, 30-day readmissions: 31.8% vs 29.3%, +Δ2.5% (95% CI – 4.2% to 9.2%), did not reach statistical significance. Frail patients vs non-frail, median (IQR) readmission costs: $35 732 ($26 638–$56 440) vs $29 319 ($22 314–$39 513) |
| Woldu et al., 2018 [29] | Retrospective (monocentric study) | 346 | ORC, RARC | 11-item mFIb | 90-day readmission | mFI: AUC 0.547 (95% CI 0.479–0.616), P = 0.2 |
| Chappidi et al., 2016 [26] | Retrospective (ACS-NSQIP Database) | 2679 | RC | 11-item mFIb | 30-day readmission | Pre-frail and frail patients (mFI ≥2) vs other patients (mFI <2): 20.3% vs 21.1%, P = 0.7 |
| Lascano et al., 2015 [21] | Retrospective (ACS-NSQIP Database) | 41,681 (5709, 7791 1443, 23,350,3388) | PN, RN, RNU, RP, ORC | 15-point mFI | Readmission rates | Frail patients (15-point mFI ≥0.20) vs non frail (15-point mFI 0–0.05): 15.9% vs 19.7%, P = 0.3 |
| Burg et al., 2018 [31] | Prospective (monocentric study) | 123, aged ≥65 years | ORC, RARC | FFC | 30/90-day readmission rates | No preoperative assessment variables were significant for 30-day readmissions. Decreased gait speed was significantly associated with 90-day readmission (P = 0.02). |
| Early mortality rates | ||||||
| Palumbo et al., 2019 [22] | Retrospective (NIS database) | 23,967 | ORC, LRC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | In-hospital mortality | Frail patients vs non-frail: 2.4% vs 1.5%, P < 0.001 Frail patients vs non-frail: OR 1.45 (95% CI 1.17–1.80), P = 0.001 |
| Michel et al., 2019 [23] | Retrospective (NDR) | 9459 | ORC, RARC | Johns Hopkins ACG frailty-defining diagnoses indicator | In-hospital mortality | Frail patients vs non-frail: 4.2% vs 1.5%, +Δ2.6%; 95% CI [0.1%-5.2%], p = 0.04 Frail patients vs non-frail: OR = 2.30, 95% CI (1.08–4.92), p = 0.03 |
| Meng et al., 2018 [25] | Retrospective (ACS-NSQIP Database) | 1516 | RC | 11-item mFIa | 30-day mortality | mFI: AUC 0.537 (95% CI 0.453–0.621), did not reach statistical significance |
| Chappidi et al., 2016 [26] | Retrospective (ACS-NSQIP Database) | 2679 | RC | 11-item mFIb | 30-day mortality | Frail and pre-frail patients (mFI ≥2) vs other patients (mFI <2): 3.5% vs 1.8%, P = 0.01 Pre-frail patients (mFI = 2) vs non-frail (mFI = 0): OR 1.24 (95% CI 0.62–2.45), P = 0.6 Frail patients (mFI ≥3) vs non-frail: OR 2.07 (95% CI 0.78–5.49), P = 0.1 |
| Taylor et al., 2019 [28] | Retrospective (ACS-NSQIP Database) | 92,999 (3823, 8154, 14,668, 2817, 13,953, 5678, 9466) | MIRP, RRP, MIRN, ORN, MIPN, OPN, RC |
5-item sFId | 30-day mortality | In RC patients: sFI: sFI = 0, 0.8%; sFI = 1, 1.5%; sFI = 2, 2.2%; sFI ≥3, 3.9%; P < 0.001 |
| Lascano et al., 2015 [21] | Retrospective (ACS-NSQIP Database) | 41,681 (5709, 7791 1443, 23,350, 3388) | PN, RN, RNU, RP, ORC | 15-point mFI and 11-item mFIa | 30-day mortality | In RC patients: Frail patients (15-point mFI ≥0.20) vs non frail (15-point mFI 0–0.05): 6.8% vs 2.1%, P = 0.005 15-point mFI: AUC 0.574, P < 0.001 The 15-point mFI was superior to the 11-item mFI in all the comparisons. |
FTR: failure to rescue; LRC: laparoscopic RC; MIPN: minimally invasive PN; MIRN: minimally invasive RN; MIRP: minimally invasive RP.
aThe mFI was calculated as the number of present factors divided by the total number of index factors. Patients were scored as: ‘robust’ (mFI = 0), ‘pre-frail’ (mFI 0.09–0.18), or ‘frail’ (mFI ≥0.27).
bThe mFI was calculated by scoring the number of risk factors per patient: 0, 1, 2, and ≥3.
cThe sFI was calculated by scoring the number of risk factors per patient: 0, 1, 2, and ≥3 (full score of 5).
dThe sFI was calculated by scoring the number of risk factors per patient: 0, 1, 2, and ≥3 (full score of 6).
eMajor complications as described by the ACS-NSQIP included coma for >24 h, stroke with residual deficits, unplanned intubation, ventilator requirement for >48 h, deep incisional surgical site infection, organ space surgical site infection, wound disruption, sepsis, septic shock, acute renal failure, progressive renal insufficiency, myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation, deep venous thrombosis, and pulmonary embolism.