Laparoscopic and robotic surgery in the elderly—is there a difference in outcomes?
- Open Access
- 14.01.2026
- main topic
Summary
Introduction
With an ageing population, the number of elderly patients undergoing surgery increases. The “older” population usually includes all patients over 65; however, further discernment of chronological age often also includes a definition of > 80 years (“octogenarians”) or > 90 years (“nonagenarians”). Age alone, however, is a poor predictor of outcomes, as physical reserves may vary significantly between individuals. Thus, the concepts of frailty and performance status are more suitable for defining which patients are at a higher risk of complications [1]. The gradual implementation of minimally invasive (MIS)—primarily laparoscopic (LS)—and, lately, robot-assisted surgery (RAS) over the past decades means that these methods can now be offered for nearly all indications, often with clear benefits in terms of operative trauma, blood loss, and length of stay (LOS) in the general population. The issues addressed by this review are (1) how MIS compares with open surgery specifically in older adults and (2) what evidence exists for RAS in these cohorts.
Methods
For this narrative review, we performed a comprehensive literature search of PubMed. We focused on meta-analyses and propensity score-matched cohorts comparing outcomes in the study population. No new statistics were performed.
Anzeige
Results
What does elderly actually mean?
The definition “elderly” in surgical research usually describes patients over the age of 65. Some studies further divide into youngest-old (65 to 74), middle-old or older-old (75–84), and oldest-old (≥ 85), while others specifically target those over 80 (“octogenarians”) [2]. Interestingly a ACS-NSQIP (American College of Surgeons – National Surgical Quality Improvement Program) analysis of over 10,000 major elective visceral surgery patients derived the threshold for worse postoperative outcomes at an age of 52.595 years [3]. However, chronological age often differs from physiological age, which may be more important for outcomes. Thus, other tools are often used to quantify physiological function. Most widely used are the World Health Organization (WHO) performance status scale first published by the Eastern Cooperative Oncology Group (ECOG) and the Karnofsky index (Table 1), both classify performance in daily life based on self-reported abilities [4]. These scores are very simple to obtain and therefore widely used, however they rely on physician and patient reports and may not fully capture geriatric vulnerabilities.
Table 1
Definitions and ranges of the Karnofsky and Eastern Cooperative Oncology Group (ECOG)/World Health Organization (WHO) scores [5]
Karnofsky score | ECOG/WHO grade | Definition |
|---|---|---|
90–100 | 0 | Fully active, able to carry on all pre-disease performance without restriction |
70–80 | 1 | Restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary nature, e.g., light housework, office work |
50–60 | 2 | Ambulatory and capable of all selfcare but unable to carry out any work activities; up and about more than 50% of waking hours |
30–40 | 3 | Capable of only limited selfcare; confined to bed or chair more than 50% of waking hours |
10–20 | 4 | Completely disabled; cannot carry out any selfcare; totally confined to bed or chair |
0 | 5 | Dead |
Frailty describes a state of reduced physiological reserve in multiple organ systems, although no strict definition exists. It can be measured with various tools, like the risk analysis index or the clinical frailty scale [1, 6‐9].
A 2016 meta-analysis including 23 studies described increased 30-day (odds ratio (OR) 1.4–8.33) and 1‑year (OR 1.1–4.97) mortality as well as a higher rate of postoperative complications (OR 1.5–4.8) and length of stay in frail patients [1]. Recent observational data from the United Kingdom found similar results, with increased LOS as well as higher odds of delirium, morbidity, and mortality at ≥ 1 year when frailty was present, even in the absence of multimorbidity [7]. This emphasizes the importance of preoperative decision-making in surgery based on more complex factors than age alone.
Surgical stress response and operative trauma—the biological case for MIS
A variety of organ systems are placed under surgical stress due to the invasive nature of surgery. Adrenal stimulation leads to raised cortisol levels mitigating immunomodulation, increased insulin resistance, and increased glucose levels [10, 11]. The resulting hyperglycemia is a significant risk factor for surgical site infection [12]. Immunomodulation may lead to reduced natural killer cell toxicity and T‑cell response [13‐15]. Catabolic changes in postoperative metabolism are associated with proteolysis, especially in muscle tissue [16]. Patient comorbidities as well as the chosen method of anesthesia may further modulate the surgical stress response [10]. In this context, reduction of surgical invasiveness can reduce the extent and impact of surgical stress on the patient. Minimally invasive surgery measurably reduces this trauma, with lower postoperative cortisol levels, a lower grade of leucodepletion, and a reduced degree of immunosuppression [17‐24].
Anzeige
It is to be expected that any surgical approach reducing operative trauma (and subsequently surgical stress) should especially be entertained in the frail and elderly population with reduced physiological reserves. Much more so as laparotomy is associated with higher rates of long-term complications such as adhesive small bowel obstruction and incisional hernia [25‐27].
However, there are several caveats for minimally invasive surgery: raised intraabdominal pressure, duration of the operation, and intraoperative positioning may put additional strain on patient physiology. Carbon dioxide insufflation and the voluntary increased intraabdominal pressure can decrease venous return, thus altering cardiac output; impair ventilation; and increase blood CO2 levels [28‐30]. Head down (Trendelenburg) positions put further stress on lung capacity and ventilation pressures, altering functional lung capacities well beyond surgery in patients with pulmonary diseases [31]. Additionally, steep Trendelenburg has been associated with raised intraocular pressure, which may be critical in elderly patients [32]. Longer operative times, often observed in MIS procedures, may be accompanied by higher cumulative doses of anesthetics, a risk of hypothermia and pressure damage to soft tissue/nerves (especially in underweight frail patients), and postoperative delirium [33, 34]. A recent multicenter study observed significantly higher rates of postoperative pulmonary complications after RAS compared to LS (19.0% vs. 9.5%; p < 0.001) [35]. However, the only independent risk factor associated with postoperative pulmonary complications was the duration of ventilation (adjusted odds ratio 1.49; 95% confidence interval (CI) 1.33–1.66; p < 0.001).
Preoperative assessment of frailty therefore plays an important role when considering (minimally invasive) surgery in elderly patients. Prehabilitation describes the concept of preoperative optimization, especially via nutrition and exercise, which has been shown to significantly improve postoperative outcomes [36]. During the operation, the lowest possible insufflation pressures, shallower Trendelenburg (if needed at all), and meticulous padding combined with lung-protective ventilation should be performed [37‐39]. Exceedingly long operative times just for the sake of MIS should be avoided and willingness to convert (if progress is missing) fostered.
What does the literature say? Evidence on MIS in select visceral surgery indications
Colorectal surgery
The colorectal cancer incidence peaks in the age group between 65 and 75 years [40]. Luo et al. performed a systematic review of 24 studies including six randomized controlled trials (RCT) [41]. The analyzed studies included patients older than 65 years and reported favorable results for minimally invasive surgery. Outcomes did not differ significantly in terms of long-term overall survival (hazard ratio (HR) 0.96, 95%CI 0.89–1.04), disease-free survival (HR 1.02, 95%CI 0.93–1.13), risk of recurrence (relative risk (RR) 1.44, 95%CI 0.90–2.30), or readmission rates (RR 1.11, 95%CI 0.88–1.40). Operative time was longer in the LS group (Weighted Mean Difference (WMD) 30.37 min, 95%CI 17.75–43.0), while blood loss was lower (WMD −78.85 ml, 95%CI −101.96–−55.75) in those undergoing laparoscopic surgery. Minimally invasive surgery was associated with shorter LOS (WMD −2.53 d, 95%CI −3.11–−1.95), quicker return of bowel function (WMD −1.06 d, 95%CI −1.20–−0.93), and a reduced pooled risk of complications (RR 0.66, 95%CI 0.60–0.74). Of note, laparoscopy was also associated with lower 90-day mortality (RR 0.70, 95%CI 0.53–0.94). In a single-center cohort of 967 stage I–III colorectal patients aged over 75 years, the authors included 662 open and 305 laparoscopically operated patients [42]. In this report, LOS was shorter in the LS group (10.3 ± 8.5 vs. 13.5 ± 9.4 days; p < 0.001), with comparable morbidity and mortality rates (p = 0.354 and 0.082, respectively). They reported a 1.97% conversion rate. However, there were significantly more T3 and T4 cancers operated in the open group (T3: 67.5% vs. 64.9%, T4: 14.5% vs. 9.2%; p = 0.009). Over a mean follow-up of 47.9 months, long-term outcomes (overal survival (OS), cancer-specific survival, and cumulative recurrence rate) did not differ between open surgery and LS. While there were significant differences between groups, this study proves the principle of MIS as a worthy option even in the older-elderly. Chok et al. analyzed ACS-NSQIP data from an even older collective in Singapore [43]. In 192 octogenarians (114 MIS) undergoing colorectal surgery, the authors observed no significant differences in baseline characteristics, postoperative complications, or 30-day and 1‑year mortality rates. While laparoscopic surgery took a mean of 52.5 min longer (232.5 vs. 180.0 min; p < 0.001), it also was associated with a reduced LOS (median 6 vs. 9 days; p < 0.001) and a significantly reduced overall cost (Singapore dollar S$ 25,583.44 vs. S$ 28,970.85; p = 0.012) because of LOS reduction. Similar results have been observed in several other publications, underscoring that LS in colorectal disease provides faster recovery with adequate oncological results also in an elderly population [44‐47].
In their 2025 publication, Niemeläinen et al. analyzed survival rates in patients over 80 years after colon cancer surgery with a special emphasis on frailty [48]. Of 227 patients included, 58 had a clinical frailty score (CFS) ≥ 5 equating to significant frailty. After a median follow-up of 4.0 years (91 days to 5 years), they observed significantly lower survival rates with increased frailty; e.g., non-frail patients (CFS 1–2) had 1‑year, 3‑year, and total follow-up survival of 100%, 89%, and 80%, respectively, as opposed to 86%, 57%, and 43%, respectively, in the highly frail population (CFS 5–9; p < 0.001). In univariate Cox regression, open surgery was associated with worse OS (HR 1.73, 95%CI 1.03–2.91; p = 0.039). These findings underscore the fact that frailty rather than chronological age is the decisive factor for postoperative outcome.
With more systems available worldwide, robotic surgery is increasing. There is already a substantial body of evidence proving oncological adequacy, further reduction of surgical trauma, and reduced LOS after RAS. However, comparative evidence for its use in elderly and frail patients is still scarce. Yang et al. retrospectively compared 95 patients aged over 60 years who underwent total mesorectal excision (TME) for rectal cancer by either RAS (45 patients) or laparoscopy (50 patients) [49]. No significant differences in baseline characteristics, conversion rate, or postoperative complications were observed. However, RAS-TME had shorter operation duration (145 min [IQR 125–187.5] vs. 180 [IQR 148.75–206.25]; p = 0.005) and a quicker restart of oral feeding (4 vs. 3 d; p = 0.048), while LOS did not differ significantly. A propensity score-matched (PSM) analysis of 5673 patients over 65 years from the U.S. National Cancer Database compared outcomes after rectal cancer resection by means of open surgery, LS, and RAS [50]. When compared to open surgery, RAS had increased odds for retrieval of 12 or more lymph nodes (p = 0.0041) and negative circumferential margins (p < 0.0001) as well as decreased odds for 90-day mortality (p = 0.0201); LS showed significantly decreased odds of 90-day mortality (p = 0.0345) when compared to open surgery, but no other significant differences were observed. Comparison between LS and RAS showed a significantly higher lymph node yield (p < 0.0249) and increased odds of negative circumferential resection margins (0.0078) in robotic surgery. Five-year OS in the open, laparoscopic, and robotic cohorts was 59.9%, 53.9%, and 63.7%, respectively, with a significant log rank (p < 0.0232) in the Kaplan–Meier estimation. An Italian study comparing RAS to LS for colonic cancer resections in the elderly observed longer operation times in RAS (right hemicolectomy [RHC] mean time 238.5 vs. 183.5 min; p = 0.004; left hemicolectomy [LHC] mean time 249.6 vs. 211.7 min; p = 0.003) but quicker recovery of bowel function (LHC 2.6 vs. 3.6; p = 0.004; RHC no significant difference) [51]. Mean LOS was significantly reduced in LHC (4.2 vs. 5.8; p = 0.004) as well as rectal and rectosigmoid resections (5 vs. 7.1; p = 0.003 and 3.7 vs. 6.2; p = 0.003) but not different in right-sided resections. Complications and oncological adequacy did not differ between groups. Xue et al. reported results of 111 patients over 80 years undergoing LS or RAS colorectal resection and found no statistically significant differences except for lower blood loss in the robotic group (76.9 ml ±39.7 vs. 161.6 ±207.1; p = 0.025) [52].
In conclusion, current data demonstrate the feasibility and oncological adequacy of RAS for colorectal surgery of the elderly, although benefits may need to be weighed against financial necessities.
Hepatobiliary surgery
Cholecystectomy was one of the earliest indications to nearly fully transition to MIS in both the elective and the emergency setting. A 2014 meta-analysis of 13 articles (two randomized and 11 observational trials) included data of over 100,000 patients over 65 years undergoing open vs. laparoscopic cholecystectomy [53]. Laparoscopy outperformed open surgery in all relevant items (morbidity 11.5% vs. 21.3%, OR 0.44, 95%CI 0.33–0.59; p < 0.00001; mortality 1.0% vs. 4.4%, OR 0.24, 95%CI 0.17–0.35; p < 0.00001; cardiac complications 0.6% vs. 1.2%, OR 0.55, 95%CI 0.38–0.80; p = 0.002; respiratory complications 2.8% vs. 5.0%, OR 0.55, 95%CI 0.51–0.60; p< 0.00001). However, due to poor methodological quality and varying inclusion/exclusion criteria, the authors concluded that these data did not favor one method over the other in the routine setting. A more recent report from Korea included 352 octogenarians and 41 nonagenarians (90 years+) with a laparoscopy rate of over 90% (96.8% and 92.7%, respectively) [54]. This seems to illustrate the reality that laparoscopy has become the standard approach in cholecystectomy. The authors observed mortality rates of 2.3% and 2.4%, respectively, and higher morbidity in the older patient group (14.7% vs. 22.0%; p < 0.001). Interestingly, mean LOS was quite high in both groups, with significantly longer hospital stays in nonagenarians (13 ± 7.22 vs. 14 ± 5.86 days; p < 0.001). These findings also support the 2020 World Society of Emergency Surgery guidelines which state, “Early laparoscopic cholecystectomy (ELC) should be the standard of care whenever possible, even in subgroups of patients who are considered fragile, such as the elderly” [55].
Two recent studies on robotic cholecystectomy (r-CHE) found conflicting results. Maegawa et al. reported from NSQIP data on 59,000 patients that r‑CHE can be performed safely, with adequate results as compared to laparoscopic surgery [56]. Mullens and colleagues, on the other hand, observed a significantly higher rate of bile duct injuries (BDI) after r‑CHE (mean rate 0.72, 95%CI 0.55–0.89 vs. 0.23, 95%CI 0.21–0.25; RR 3.12; 95%CI 2.34–3.91), with a higher risk for reoperation in the robot-assisted compared to the laparoscopic group (RR 1.47; 95%CI 1.35–1.59) [30]. One limiting factor of this study was its retrospective analysis of Medicare billing data and the risk of coding errors. A possible explanation for higher rates of BDI in r‑CHE may be that cholecystectomy is considered an “easy” training operation to adopt a robotic platform [57]. Concerning the use of r‑CHE in the elderly and frail, no data exist to our knowledge.
There is a growing body of evidence supporting laparoscopic liver resection (LLR) in older patients. A multicenter analysis comparing LLR to open surgery propensity score matched 52 octogenarians per group [58]. There was no significant difference in operative time (269 min [94–698] vs. 261 [88–747]; p = 0.68) or duration of the Pringle maneuver (59 min [0–120] vs. 66 [0–209]; p = 0.11). Blood loss (239 ml [15–2750] vs. 55 [10–4820]; p < 0.01) and LOS (16 days [11–40] vs. 14 [7–59]; p < 0.01) were significantly lower in the laparoscopic group, as was the incidence of cardiopulmonary complications (5 [9.6] vs. 0 [0]; p < 0.01). A study on surgery of liver tumors > 5 cm in patients over 70 years propensity score matched two groups (90 per group) for age, BMI, surgical center, underlying liver cirrhosis, comorbidities, extent of the resection, and tumor size [59]. The authors reported shorter LOS (7 vs. 9 days; p = 0.01) and a lower rate of R1 resections (4.4% vs. 13.3%; p = 0.03) in LLR. No significant differences were observed for operative time, need for blood transfusion, severe postoperative complications, post-hepatectomy liver failure, or in-hospital mortality. Long-term results showed no significant difference in OS and DFS after a median follow up of 35 months (95%CI 27.6–42.3). Similar results were observed by Verhoeff et al. after propensity score matching their cohort of 2674 patients: reduced LOS (−1.99 days; p < 0.001), while all other relevant factors were comparable to open surgery [60].
A single-center PSM analysis from Denmark reported the short-term results after open vs. robotic liver resection (RLR) in patients ≥ 70 years [61]. Mean operative time was significantly longer in RAS (165.1 ± 55.0 vs. 256.6 ± 112.6 min; p < 0,001). Mean tumor diameter was higher in the robotic group (30.9 ± 16.6 vs. 42.8 ± 32.8 mm; p = 0.040), while blood loss was significantly lower (821.2 ± 719.4 vs. 155.2 ± 146.3 mL; p < 0.001). Robotic liver resection was associated with significantly lower complication rates (0% vs. 16.1%; p = 0.022) and a shorter LOS (3.4 vs. 6.5 days; p = 0.006). These effects were more pronounced in minor liver resections. Similarly, a multicenter study with matched cohorts of 106 patients in open and robotic liver surgery observed longer operation times (median [IQR], 295 [190–370] minutes vs. 200 [165–255] minutes; p < 0.001) but reduced LOS (median [IQR] 4 [3–6] days vs. 10 [7–13] days; p < 0.001) and lower rates of ICU admission (7 [6.6%] vs. 21 [19.8%]; p = 0.002) and post-hepatectomy liver failure (8 [7.5%] vs. 30 [28.3%]; p= 0.001) after RLS [62]. Of note, this study included patients of all age groups; after PSM, however, the median age was 69 (IQR 63–72) and 67 (IQR 59–72) for open and robotic surgery, respectively.
Robotic liver resection showed comparable results to LLR in a recent publication. In their NSQIP analysis of 2210 patients aged over 65 years undergoing liver surgery, Parente et al. reported longer operation times (32.62 min, 95%CI 15.82–49.41 min; p < 0.001) and shorter LOS (−0.72 days, 95%CI −1.29%–−0.16; p = 0.0124) after RLR as compared to LLR after propensity score matching [23]. Short-term complication rates and oncological outcomes did not differ significantly.
Gastric surgery
He and colleagues performed a systematic review comparing laparoscopic vs. open gastrectomy for gastric cancer in patients aged over 80 years [63]. Eight cohort studies including 807 patients were included and reemphasized the perioperative results seen in colorectal surgery: longer operation times (WMD 30.48 min, 95%CI 14.80–46.16; p < 0.001; I2 = 74%), reduced blood loss (WMD −166.96 ml, 95%CI −220.60–113.31; p < 0.001; I2 = 85%), and similar lymph node harvest in laparoscopic gastrectomy. Postoperatively, LS was associated with earlier bowel movements (WMD −0.83 days, 95%CI −1.13–0.53; p < 0.001; I2 = 51%) and a shorter LOS (WMD = 0.78 days, 95%CI 1.17–0.38; p < 0.001; I2 = 7%). While overall complications were lower in the MIS group (OR 0.54, 95%CI 0.35–0.75; p = 0.003; I2 = 0%), no significant difference was observed in major complications (Clavien–Dindo IIIa and higher). Interestingly, pooled analysis of five studies comparing 5‑year OS favored laparoscopy (OR 1.66, 95%CI 1.04–2.67; p = 0.03; I2 = 0%). Disease-specific survival was also higher in LS (three studies; OR 3.23, 95%CI 1.61–6.47; p < 0.001; I2 = 0%). A study using ACS-NSQIP data including 2661 patients undergoing gastrectomy (624 laparoscopic) reported a significantly lower rate of postoperative pneumonia (OR 0.45 [0.21, 0.98]; p = 0.044) in patients over 70 years after laparoscopic vs. open gastrectomy [64]. These data suggest that age alone is no contraindication to MIS in gastric cancer surgery.
A propensity score-matched analysis compared RAS to open gastric cancer surgery in a cohort aged over 70 years including 43 matched patients per group [65]. Garbarino et al. observed a significantly longer operative time (273.8 vs. 193.5 min; p < 0.01) in robotic surgery, with no differences in LOS, mortality, morbidity, oncological adequacy, and OS. However, major complications (6.9 vs. 16.3%, OR 2.592, 95% CI 0.623–10.785; p = 0.313) and postoperative pain (0.95 vs. 1.24; p = 0.042) were both lower in der RAS cohort. A large multicenter study from China retrospectively included 1393 patients ≥ 65 years undergoing either laparoscopic or robotic gastrectomy for cancer [66]. Confounding biases were addressed by inverse probability of treatment weighting and propensity score matching, resulting in matched groups of 363 patients each. Operative time was shorter in the laparoscopic group (216.93 vs. 205.56 min; p = 0.008), while RAS showed lower blood loss (89.36 vs. 103.39 mL; p = 0.046). No significant differences were observed in lymph node harvest, resection margin positivity, or conversion rate. Patients treated by RAS left the hospital significantly earlier (9.62 vs. 10.47 days; p = 0.017). There was no difference in complications, reoperation rate, or mortality rate. Long-term oncological outcomes did not differ.
A 2020 systematic review including 40 studies and 17,712 patients compared robotic to laparoscopic gastrectomy, without specifically addressing elderly patients [67]. While the authors observed the classical longer operation time and reduced blood loss, they also found a significantly lower rate of surgical complications (Clavien–Dindo ≥ 3; OR 0.66, 95%CI 0.49–0.88; p = 0.005) and a higher lymph node yield after RAS (MD 1.84, 95%CI 0.84–2.84; p = 0.0003).
Emergency surgery in the elderly
Minimally invasive emergency surgery in an elderly population is supported across several indications [68]. In acute cholecystitis, guidelines recommend early laparoscopic cholecystectomy also in high-risk/elderly cohorts with shorter LOS and lower complications [53‐55].
In acute appendicitis, a meta-analysis including over 330,000 patients found reduced postoperative mortality (OR 0.33; 95%CI 0.28–0.39), postoperative complications (OR 0.65, 95%CI 0.62–0.67), and wound infections (OR 0.27, 95%CI 0.22–0.32) following laparoscopy [69]. These results are reflected by a 2020 multi-society guideline recommending MIS appendectomy in elderly and frail patient groups due to “reduced LOS, morbidity and costs” [70].
Perforated peptic ulcer disease (PUD) is another indication where laparoscopic repair in stable patients is supported by the literature in terms of a reduced wound infection rate and LOS paired with comparable mortality to open repair [71‐73]. Data specifically regarding elderly patients are scarce. An Italian publication demonstrated the feasibility of laparoscopic PUD repair in an elderly cohort [74]. Of 67 patients, 33 were treated by LS and showed a non-significantly lower rate of blood loss and shorter LOS combined with similar postoperative morbidity and mortality.
Apart from cholecystectomy data there are very few data on robot-assisted surgery in elderly emergencies. Indications for RAS in the acute setting must therefore be decided on an individual basis, based on patient- as well as on surgeon-related and institutional factors. While evidence for RAS is not yet available, laparoscopy remains the standard MIS platform for several select emergency surgery indications in older adults.
Conclusion
This narrative review gives a short overview of the available evidence for minimally invasive and robotic surgery in the elderly. While advanced age can be associated with worse outcomes, frailty predicts adverse events with higher significance. In general, minimally invasive surgery should also be the go-to approach in elderly patients, when technically feasible in an adequate time and oncologically adequate. It reduces morbidity and the length of hospital stay and decreases surgical trauma. The growing use of robot-assisted surgery is reflected in the literature, but data—while results are promising—on its use in elderly patients are still limited, and no definitive conclusions can be drawn.
Conflict of interest
V. Frieders-Justin, B. Glaser, and P. Stiegler declare that they have no competing interests. S. Uranues is a member of the editorial board of european surgery and was recused from the handling of this manuscript.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.