Colorectal cancer (CRC) screening programs produce risks, including those derived from colorectal surgeries. The objective of this analysis is to evaluate the complications associated with the surgery.
Patients and methodsRetrospective analysis including patients who required colorectal surgery within the population-based CRC screening program in Galicia (May 2013–June 2019). We analyzed the indication for surgery and the rate of in-hospital (mild I-II, severe III-V, Clavien-Dindo classification) and at discharge complications. We performed a multivariate analysis to determine the variables independently associated.
ResultsIn the analyzed period, 1092 patients underwent surgery (benign lesion 16.5%, pT1 CRC 18.2%, rest of CRC 64.6%) laparoscopic approach in 69.8% of the cases. In-hospital complications were detected in 19.2% of patients (mild: 13.4%, severe: 5.9%, deaths: 0.2%) and at discharge in 159 (14.6%) patients. Male sex was associated with in-hospital complications (OR 2.0 95% CI 1.3−3.0). The variables associated with severe complications were: male sex (OR 2.6, 95% CI 1.2−5.5), tertiary hospital (OR 0.5, 95% CI (0.2−0.9) and ECOG I (OR 0.2, 95% CI 0.05−0.6). The factors associated with complications after discharge were age ≥60 years (OR 1.5, 95% CI 1.0–2.3), rectal location (OR 1.6, 95% CI 1.1–2.3) and in-hospital complications (OR 2.2, 95% CI 1.5–3.2).
ConclusionsSurgery is the main cause of morbidity and mortality associated with a CRC screening program. These results must be taken into account in the decision making of lesions that are candidates for endoscopic resection.
Los programas de cribado de cáncer colorrectal (CCR) producen riesgos, entre ellos los derivados de las cirugías colorrectales. El objetivo de este análisis es evaluar las complicaciones asociadas a la cirugía.
Pacientes y métodosAnálisis retrospectivo de los pacientes que requirieron cirugía colorrectal dentro del programa poblacional de cribado de CCR de Galicia (mayo de 2013-junio de 2019). Analizamos laindicación de la cirugía y la tasa de complicaciones intrahospitalarias (leves I-II, graves III-V, clasificación Clavien–Dindo) y al alta. Determinamos mediante un análisis multivariante las variables asociadas a su aparición.
ResultadosEn el periodo analizado, 1092 pacientes fueron intervenidos (lesión benigna 16.5%, CCR pT1 18.2%, resto CCR 64.6%), por vía laparoscópica en el 69.8% de los casos. Se detectaron complicaciones intrahospitalarias en el 19.2% de los pacientes (leves: 13.4%, graves: 5.9%, fallecimientos: 0.2%) y al alta en 159 (14.6%) pacientes. El sexo masculino se asoció a las complicaciones intrahospitalarias (OR 2.0 IC 95% 1.3−3.0). Las variables asociadas a las complicaciones graves fueron: sexo masculino (OR 2.6, IC 95% 1.2−5.5), hospital terciario (OR 0.5, IC 95% (0.2−0.9) y ECOG I (OR 0.2, IC 95% 0.05−0.6). Los factores asociados a las complicaciones tras el alta fueron edad ≥60 años (OR 1.5, IC 95% 1.0–2.3), la ubicación rectal (OR 1.6, IC 95% 1.1–2.3) y complicaciones intrahospitalarias (OR 2.2, IC 95% 1.5–3.2).
ConclusionesLa cirugía es la principal causa de morbimortalidad asociada a un programa de cribado de CCR. Estos resultados deben ser tenidos en cuenta en la toma de decisiones en lesiones candidatas a resección endoscópica.
Colorectal cancer (CRC) is a highly prevalent disease in our society, with a known natural history that permits early detection and reduced associated mortality. For this reason, it is a condition that benefits from population screening to detect precancerous lesions or early-stages tumours.1 The effectiveness of the different CRC screening programme options (stool tests, sigmoidoscopy, colonoscopy) in reducing the mortality and incidence of CRC has been demonstrated both in clinical trials and in systematic reviews.2,3
In screening programmes, the benefit obtained must outweigh any harm, caused, for example, by overdiagnosis, overtreatment, false positives, false security, incidental findings or complications.4 Although the complications associated with diagnostic tests are well described in CRC screening,1,5 there is no such certainty regarding overdiagnosis and overtreatment. Overdiagnosis is defined as the diagnosis of a medical condition or illness that would not cause symptoms or death during a patient's lifetime. In the case of CRC screening, the treatment of non-invasive CRC and overdiagnosed polyps may be regarded as overtreatment.6
The endoscopic resection of colorectal polyps is the key to reducing the incidence and mortality of CRC.7 Although the side effects are limited, mainly post-polypectomy syndrome, rectal bleeding and perforation, they are responsible for the majority of colonoscopy-related complications.1 Endoscopic resection removes up to 90% of advanced complex polyps.8 However, the introduction of CRC screening programmes has increased the number of colectomies for benign polyps. In the United States, up to 25% of colectomies are performed for non-malignant polyps.9 Related in-hospital mortality and morbidity have reached 0.8% and 25.3%, respectively.10
There is scant evidence available on the risks associated with surgery in a CRC screening programme. For this reason, we decided to conduct an observational study to evaluate the complications associated with surgery as well as the factors associated with these complications in a CRC population screening programme.
Material and methodsStudy designCross-sectional retrospective study in which all patients who required colorectal surgery after a colonoscopy performed within the CRC population screening programme as of its launch in May 2013 until June 2019 were included.
Description of the CRC population screening programme in GaliciaThe CRC population screening programme in Galicia is based on the performance of a biennial immunological test for the detection of faecal occult blood (FOB) with a cut-off point of 20 μg of haemoglobin/g of faeces in subjects aged between 50 and 69 years. Between its launch in June 2013 and until July 2019, 721,349 people were invited to participate in the screening programme. The programme was started in the health areas of Ferrol in 2013, Ourense in 2015, Pontevedra, Santiago and Lugo in 2016, and A Coruña and Vigo in 2017. The screening programme is coordinated by the Department of Public Health of the Regional Ministry of Health, which is in charge of identifying the subjects, inviting them to participate, receiving the results of the FOB test and summoning patients with a positive result for follow-up.11
DatabasePatients with colorectal surgery were identified in the information system of the CRC screening population programme in Galicia. Surgery was confirmed by consulting the unified electronic medical record (IANUS) of the Galician Health Service.
Variables collectedData were collected on demographics (age and sex), the result of the FOB test, anaesthetic risk and quality of life data measured by the American Society of Anesthesiologists(ASA) and Eastern Cooperative Oncology Group (ECOG) scales and the level of specialisation at the hospital where the patient underwent surgery. The following surgery-specific data were identified: type of surgery, location of the lesion, neoadjuvant treatment performed, surgical approach, duration of admission, complications during admission and after discharge and survival at the time of the analysis. To classify the severity of complications during the hospital stay, we used the Clavien–Dindo classification,12 considering those classified as grades I-II as mild, those requiring reoperation as grade III, those requiring admission to a critical care unit as grade IV and patient deaths as grade V. Complications were defined as minor for grades I-II and major for grades III-V.
Statistical analysisFirst, a descriptive analysis of the subjects included was performed. We used the median and interquartile range for quantitative variables and the absolute number and frequency for qualitative variables. We determined which variables were associated with the appearance of both overall and major complications during admission and after discharge. Initially, we carried out a univariate analysis using the chi-square test for qualitative variables and Student's t-test for quantitative variables. Statistically significant or clinically relevant variables were included in a logistic regression. Associations were expressed as odds ratio (OR) with the 95% confidence interval (CI). The statistical analysis was performed using the statistical package IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY, USA; IBM Corp.
Ethical considerationsThe study was authorised by the Clinical Research Ethics Committee of Galicia (2018/593). Since the study was based on a database analysis, informed consent was not required. The information was accessed in accordance with current Spanish and European legislation.
ResultsDescription of the sample and surgeriesIn the period analysed, we identified 1,092 patients participating in the CRC screening population programme who required surgical treatment. Only three patients underwent surgery outside the Galician Health Service and for whom no surgery-related information could be retrieved. The epidemiological data of the patients included in the study are shown in Table 1. The surgical patients were predominantly male (66.3%) aged over 60 years (72%), with a faecal haemoglobin concentration higher than 200 μg/g of faeces in 42.4% of patients, a low pre-anaesthetic risk (ASA I) in 60.8%, and preserved quality of life (ECOG 0) in 82.0%.
Baseline characteristics of operated patients participating in colon cancer screening.
| Frequency | Percentage (%) | |
|---|---|---|
| Sex (n = 1,092) | ||
| Female | 368 | 33.7 |
| Male | 724 | 66.3 |
| Age (n = 1,092) | ||
| <60 years | 306 | 28.0 |
| ≥60 years | 786 | 72.0 |
| Faecal haemoglobin (n = 1,012) | ||
| <100 μg Hb/g | 306 | 30.2 |
| 100−199 μg Hb/g | 277 | 27.4 |
| ≥200 μg/g | 429 | 42.4 |
| ASA (n = 1,005) | ||
| I | 611 | 60.8 |
| II | 343 | 34.1 |
| III | 51 | 5.1 |
| ECOG (n = 1,005) | ||
| 0 | 824 | 82.0 |
| 1 | 181 | 18.0 |
| Colonoscopies performed (n = 969) | ||
| 0 | 122 | 12.6 |
| 1 | 807 | 83.3 |
| >1 | 40 | 4.1 |
| Tertiary hospital (n = 1,092) | ||
| No | 714 | 65.4 |
| Yes | 378 | 34.6 |
| Health area (n = 1,092) | ||
| Ferrol | 236 | 21.6 |
| Ourense | 232 | 21.2 |
| Pontevedra | 113 | 10.3 |
| Lugo | 133 | 12.2 |
| Santiago | 175 | 16.0 |
| Vigo | 95 | 8.7 |
| Coruña | 108 | 9.9 |
ASA, American Society of Anesthesiologists; ECOG, Eastern Cooperative Oncology Group.
As shown in Table 2, surgery was performed in relation to invasive cancer in 82.9% of the cases, while the rest were due to benign causes or complications associated with the endoscopic procedure. The resected lesion was located in the rectum in 30.3% of the patients, and 15.8% (172) of the patients required pre-operative neoadjuvant treatment. Overall, the laparoscopy was the most commonly used approach, in up to 69.7% of the patients. The most common types of surgery were sigmoidectomy (28.7%), right hemicolectomy (24.9%) and low anterior resection (23.1%). The median admission time was seven days (interquartile range 6–10). Post-surgical in-hospital complications were detected in 19.2% of the surgical patients. They were mild (I-II) in 13.4% of the patients and serious (III-V) in 5.9%, with two deaths (0.2%).
Description of the lesion, surgical procedure and associated complications in the study patients.
| n (%) | |
|---|---|
| Type of lesion (n = 1,092) | |
| Benign | 180 (16.5%) |
| Complications | 5 (0.5%) |
| CRC pT1 | 199 (18.2%) |
| Other CRCs | 705 (64.6%) |
| Other | 3 (0.3%) |
| Surgical approach | |
| Transanal | 62 (5.7%) |
| Laparoscopy | 689 (63.1%) |
| Converted laparoscopy | 73 (6.7%) |
| Open surgery | 264 (24.3%) |
| Type of surgery | |
| Abdominoperineal amputation | 19 (1.7%) |
| Extended appendectomy | 10 (0.9%) |
| Segmental colectomy | 2 (0.2%) |
| Subtotal colectomy | 34 (3.1%) |
| Right hemicolectomy | 272 (24.9%) |
| Left hemicolectomy | 82 (7.5%) |
| Low anterior resection | 252 (23.1%) |
| Segmental resection | 37 (3.4%) |
| Transanal resection | 60 (5.5%) |
| Sigmoidectomy | 313 (28.7%) |
| Other | 10 (0.9%) |
| Classification of complications (Clavien–Dindo) | |
| 0 | 881 (80.8%) |
| I | 89 (8.2%) |
| II | 57 (5.2%) |
| III | 34 (3.1%) |
| IV | 28 (2.6%) |
| V | 2 (0.2%) |
| Complications at discharge | |
| Yes | 159 (14.6%) |
| No | 930 (85.2%) |
| Lost | 3 (0.3%) |
| Type of complications at discharge | |
| Occlusive/subocclusive symptoms | 32 (3.0%) |
| Anastomotic stricture | 11 (1.0%) |
| Hernia/eventration | 31 (2.9%) |
| Surgical wound infection | 8 (0.7%) |
| Haematoma/seroma | 9 (0.8%) |
| Ostomy complications | 7 (0.6%) |
| Intra-abdominal collections | 23 (2.1%) |
| Secondary symptoms | 15 (1.4%) |
| Systemic complications | 8 (0.7%) |
| Rectal bleeding | 6 (0.6%) |
| Final status | |
| Alive | 1,051 (96.2%) |
| Dead | 41 (3.8%) |
CRC, colorectal cancer.
After a median follow-up of 25 months (interquartile range 17–34), postoperative complications were detected in 159 patients (14.6%), predominantly related to occlusive or subocclusive conditions (n = 32; 3.0%), hernias or eventration (n = 31; 2.9%) and intra-abdominal collections (n = 23; 2.1%). Forty-one (41) patients died during follow-up (3.8%).
Factors associated with the appearance of complicationsTable 3 shows the variables in the univariate analysis associated with the appearance of in-hospital and post-discharge complications, both overall and serious. In this analysis, in-hospital complications were statistically significantly associated with the type of surgery and the type of hospital. Serious complications were statistically associated with the type of approach, the hospital and the performance status. Finally, post-discharge complications were associated with age, surgical approach and the appearance of in-hospital complications.
Variables associated with complications in the univariate analysis.
| Evaluated variables | Complications (n = 210) | pa | Serious complications (n = 64) | pa | Complications after discharge (n = 159) | pa |
|---|---|---|---|---|---|---|
| Sex | <0.001 | 0.04 | 0.2 | |||
| Female | 48 (13.1%) | 14 (3.8%) | 48 (13.0%) | |||
| Male | 162 (22.4%) | 50 (6.9%) | 111 (15.4%) | |||
| Age | 0.5 | 0.7 | 0.04 | |||
| <60 years | 55 (18%) | 19 (6.2%) | 34 (11.1%) | |||
| >60 years | 155 (19.7%) | 45 (5.7%) | 125 (16.0%) | |||
| FOB | 0.1 | 0.5 | 0.06 | |||
| <100 μg Hb/g | 67 (21.9%) | 18 (5.9%) | 38 (12.5%) | |||
| 100−199 μg Hb/g | 50 (18.1%) | 17 (6.1%) | 51 (18.5%) | |||
| ≥200 μg Hb/g | 70 (16.4%) | 19 (4.4%) | 56 (13.1%) | |||
| ASA | 0.1 | 0.2 | 0.4 | |||
| ASA I | 99 (16.2%) | 33 (5.4%) | 93 (15.2%) | |||
| ASA II | 71 (20.7%) | 15 (4.4%) | 45 (13.2%) | |||
| ASA III | 11 (21.6%) | 5 (9.8%) | 5 (10.0%) | |||
| ECOG | 0.7 | 0.04 | 0.3 | |||
| 0 | 147 (17.9%) | 49 (6.0%) | 121 (14.7%) | |||
| 1 | 34 (18.8%) | 4 (2.2%) | 22 (12.2%) | |||
| Tertiary hospital | 0.004 | 0.01 | 0.1 | |||
| No | 155 (21.7%) | 51 (7.2%) | 112 (15.7%) | |||
| Yes | 55 (14.6%) | 13 (3.4%) | 47 (12.5%) | |||
| Lesion location | 0.09 | 0.06 | 0.02 | |||
| Rectum | 74 (22.4%) | 28 (7.9%) | 63 (19.0%) | |||
| Rest of colon | 136 (17.9%) | 38 (5.0%) | 96 (12.6%) | |||
| Neoadjuvant | 0.4 | 0.05 | 0.1 | |||
| No | 173 (18.8%) | 48 (5.2%) | 128 (14.0%) | |||
| Yes | 37 (21.5%) | 16 (9.3%) | 31 (18.0%) | |||
| Approach | 0.002 | 0.006 | <0.001 | |||
| Open surgery | 68 (25.8%) | 25 (9.5%) | 60 (23.0%) | |||
| Laparoscopic surgery | 109 (15.9%) | 31 (4.5%) | 77 (11.2%) | |||
| Converted laparoscopic surgery | 23 (31.5%) | 7 (9.6%) | 13 (17.8%) | |||
| Transanal | 10 (16.4%) | 1 (1.6%) | 7 (11.5%) | |||
| Hospital complications | <0.001 | |||||
| No | 109 (12.4%) | |||||
| Yes | 50 (24.2%) | |||||
ASA, American Society of Anesthesiologists; ECOG, Eastern Cooperative Oncology Group; FOB: faecal occult blood.
In the multivariate analysis, male sex was independently associated with in-hospital complications, both overall (OR: 2.0; 95% CI: 1.3−3.0) and serious (OR: 2.6, 95% CI: 1.2−5.5). Surgery at a tertiary hospital (OR: 0.5; 95% CI: 0.2−0.9) and ECOG 1 (OR: 0.2; 95% CI: 0.05−0.6) reduced the risk of serious complications. Regarding the factors associated with post-discharge complications, both age over 60 years (OR: 1.5; 95% CI: 1.0–2.3) and rectal location of the lesion (OR: 1.6; 95% CI: 1.1–2.3), as well as the presence of in-hospital complications (OR: 2.2; 95% CI: 1.5–3.2), were independently associated with an increased risk, as can be seen in Table 4.
Variables associated with the appearance of complications in the multivariate analysis.
| Evaluated variables | Complications during admission (n = 210) OR (95% CI)a | Serious complications during admission (n = 64) OR (95% CI)a | Complications at discharge (n = 159) OR (95% CI)a |
|---|---|---|---|
| Sex | |||
| Female | 1 | 1 | 1 |
| Male | 2.0 (1.3−3.0) | 2.6 (1.2−5.5) | 1.2 (0.8−1.8) |
| Age | |||
| <60 years | 1 | 1 | 1 |
| >60 years | 1.0 (0.7−1.4) | 0.8 (0.4−1.6) | 1.6 (1.0−2.5) |
| FOB | |||
| <100 μg Hb/g | 1 | 1 | 1 |
| 100−199 μg Hb/g | 0.7 (0.4−1.0) | 1.0 (0.5−2.0) | 1.6 (1.0−2.7) |
| ≥200 μg Hb/g | 0.6 (0.4−1.0) | 0.7 (0.3−1.4) | 1.1 (0.7−1.8) |
| ASA | |||
| ASA I | 1 | 1 | 1 |
| ASA II | 1.3 (0.9−1.9) | 1.0 (0.5−1.9) | 0.8 (0.5−1.2) |
| ASA III | 1.4 (0.6−3.2) | 5.7 (1.6−20.1) | 0.7 (0.2−1.9) |
| ECOG | |||
| 0 | 1 | 1 | 1 |
| 1 | 0.8 (0.5−1.3) | 0.2 (0.05−0.6) | 0.8 (0.5−1.5) |
| Tertiary hospital | |||
| No | 1 | 1 | 1 |
| Yes | 0.7 (0.5−1.0) | 0.5 (0.2−0.9) | 0.9 (0.6−1.3) |
| Approach | |||
| Transanal | 1 | 1 | 1 |
| Laparoscopic surgery | 0.9 (0.4−1.9) | 0.9 (0.4−1.8) | 0.9 (0.4−2.0) |
| Converted laparoscopic surgery | 2.0 (0.8−4.9) | 2.0 (0.8−4.9) | 1.7 (0.6−4.7) |
| Open surgery | 1.6 (0.8−3.5) | 1.6 (0.8−3.5) | 1.6 (0.7−3.9) |
| Complications during admission | |||
| No | 1 | ||
| Yes | 2.4 (1.6−3.6) | ||
ASA, American Society of Anesthesiologists; ECOG, Eastern Cooperative Oncology Group; CI, confidence interval; OR, odds ratio; FOB, faecal occult blood.
We have described the rates of complications, both in-hospital and at discharge, after colorectal surgeries performed within a CRC population screening programme, as well as the factors associated with them. These rates are low and are within what is expected in cohorts of young patients with little comorbidity. The information we provide is relevant to determine the risks associated with participation in a CRC population screening programme. Moreover, these data are key in making decisions about the treatment of neoplastic colorectal conditions that are candidates for endoscopic treatment: pT1 CRC and advanced benign lesions.
Population-based CRC screening targets a predominantly asymptomatic and previously healthy population and should therefore only be performed after a careful consideration of the possible harm and benefits. In fact, we must emphasise that the vast majority of the patients in our study were ECOG 0 and 1, with only 51 ASA III patients. The harmful effects of CRC screening include colonoscopy complications, possible overdiagnosis, psychological impact and treatment-associated complications. Information about the possible secondary consequences of participating in a CRC screening programme, especially those derived from post-surgical complications, is essential in deciding whether to participate.5 In this sense, there is abundant information available on the risks associated with colonoscopy, although information about overdiagnosis13 or surgery-related complications is very limited.5
In a systematic review analysing the risks associated with CRC screening, eight studies on post-surgical morbidity and mortality were identified. In-hospital complication rates ranged from 14% to 24%, and major complications from 0% to 14%. Reported mortality rates were low, ranging from 0% to 3.3%. The largest study (n = 68,306) reported a 5% reoperation rate and no 30-day mortality.14 Our results are comparable with published findings. Additionally, our study provides information about long-term complications after colorectal surgery in the context of a population screening programme. This rate, 14.6%, is clearly lower than those documented after colorectal surgeries outside population screening programmes. Similarly, the risks associated with colorectal surgery in screening programmes are lower than those described in the general population due to the characteristics of the population participating in CRC screening programmes. In a prospective study carried out in our setting, post-surgical complications in the general population were detected in up to 40% of patients during admission, in 15% in the month after discharge, and 25% in the first year after surgery.15 These differences with the data documented in our study are related to various factors associated with the risk of complications: age, comorbidity, quality of life and surgical approach.16–18 We must bear in mind that the patients in our study are a population group with little comorbidity and a good baseline situation, which explains why the percentage of postoperative complications is lower. Our study attributed statistical significance to the male sex, since this is an independent risk factor for post-surgical complications of colic anastomoses described in the literature. Open, more invasive surgeries are also known to be more associated with a higher incidence of postoperative complications. Another interesting finding from our study was that undergoing surgery in a tertiary hospital was a protective factor in the development of complications, both intraoperatively and in postoperative care, perhaps accounted for by the team's greater experience or the greater availability of resources (technical and human). In any case, new studies would be needed to elucidate this factor further.
The data provided are especially important in situations in which endoscopic treatment can be an alternative to surgery. In decision-making, not only immediate complications must be assessed, but also those that can limit the patient's quality of life in the long term.19 This is particularly important in patients with complex benign lesions, which account for 16.5% of surgeries. To reduce the risk of complications, it would be advisable to have reference centres to centralise complex endoscopic resections. Moreover, minimally-invasive approaches should be promoted if surgical treatment is eventually required.20 Another situation in which this information is relevant is in early-stage CRC (pT1), especially in invasive adenocarcinomas on endoscopically-resected adenomas. In decision-making with the patient, the risk of residual disease after endoscopic resection must be weighed up not only against the risk of in-hospital mortality but also against in-hospital and long-term complications.11
Our study has several limitations, mainly related to data collection. Our analysis is retrospective. We obtained the data collected in the screening programme information system, although we were unable to calculate other comorbidity indices or the treatments the patient was undergoing at the time of diagnosis. Neither were we able to collect information associated with the risk of the procedure, such as nutrition or preoperative anaemia, or logically the surgeon's experience. On the other hand, we were able to assess the complications in a multicentre context, since we analysed the complications in all the Galician centres participating in the population screening programme.
To conclude, we have described the indications for surgery, the approach, the type of surgery, and above all both in-hospital and long-term complications in a population-based CRC screening programme in our setting. This information is relevant in estimating the risks associated with the screening programme and in decision-making in situations in which endoscopic treatment is an alternative to surgery.
FundingThis study was supported by grants from the Academia Médico Quirúrgica de Ourense [Medical Surgical Academy of Ourense] and the Instituto de Salud Carlos III [Carlos III Health Institute] through project PI17/00837 (co-funded by the European Regional Development Fund "A way to make Europe" and the European Social Fund "Investing in your future"). Ciberehd is funded by the Instituto de Salud Carlos III.
Conflicts of interestThe authors have no conflicts of interest to declare.
