Int J Colorectal Dis (2014) 29:631–638 DOI 10.1007/s00384-014-1843-9
ORIGINAL ARTICLE
The impact of shortened postgraduate surgical training on colorectal cancer outcome A. Currie & E. M. Burns & P. Aylin & A. Darzi & O. D. Faiz & P. Ziprin & on behalf of the Imperial College Healthcare NHS trust Colorectal Cancer MDT
Accepted: 11 February 2014 / Published online: 6 March 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Background Shortened postgraduate surgical training reforms, known as Calman, have altered delivery of surgical training in the UK with reduced working hours and training time aiming to produce a more subspecialised workforce. Aims This study aims to compare rectal cancer surgical outcomes of Calman-trained consultants in a single institution to published data. Additionally, the study compared colorectal cancer surgical outcome between Calman-trained consultants (CTCs) and non-Calman consultants (NCTCs) in a national dataset. Methods Local dataset Clinicopathological outcome of rectal cancer resection undertaken by CTCs in a single institution (2006–2010) were compared against NCTC counterparts. National dataset All elective colorectal cancer resections between 2004 and 2008 in English NHS hospitals were included. CTCs (present from 2004 onwards) were compared to NCTCs (present prior to 2004). Outcome measures included 30-day in-hospital mortality, reoperation and readmission rates. Results Local dataset One hundred thirteen patients were operated under five CTC. The 30-day in-hospital mortality Presented in poster form at the Association of Surgeons of Great Britain and Ireland International Surgical Congress, Bournemouth, UK. May 2011 What’s new in this paper? This paper is the first to examine the outcomes of shortened postgraduate training methods in outcomes in the setting of predominantly surgically managed rectal cancer. A. Currie : E. M. Burns : A. Darzi : O. D. Faiz : P. Ziprin (*) Department of Surgery, Imperial College, St Mary’s Hospital, Praed Street, London W21NY, UK e-mail: [email protected] P. Aylin Dr. Foster Unit, Department of Primary Care and Social Medicine, Imperial College, 1st Floor Jarvis House, 12 Smithfield Street, London EC1A 9LA, UK
for CTCs (1 %) was favourable compared to published rates (3–5 %). Local recurrence rate (4.4 %) was comparable to NCTC (3.6 %). National dataset Between 2004 and 2008, 44,106 patients underwent elective colorectal resection. Multiple regression demonstrated CTC patients had a reduced length of stay and reduced reoperation rate. No difference in mortality and unplanned readmission rates were seen. Conclusion CTCs have similar safety outcome to NCTCs for colorectal cancer resection procedures. Further work is needed to assess the impact of further training reductions on clinical outcome. Keywords Calman . Colorectal cancer . Postgraduate surgical training
Introduction Quality assurance has become an increasingly important aspect of colorectal surgery. Technical outcomes, such as circumferential resection margin (CRM) status [1], have been suggested as potential quality measures. Significant changes have occurred to the technical training of surgeons in the UK [2–8], Australasia [9, 10] and North America [11, 12] and led to a reduction in training hours. One of the aims of such reforms is to deliver a more specialised workforce. The Calman reforms, fully instituted in 1997, overhauled UK postgraduate medical training [3] in an attempt to deliver a more subspecialised surgeon. Similar reforms took place in the European and US training systems. These changes resulted in concerns regarding reduced surgical trainee operating volume [8], shorter overall surgical training and the potential impact on patient outcome [2–7]. To date, no studies have compared outcome amongst Calman-trained consultant
632
(CTC) surgeons with non-Calman consultant (NCTC) counterparts. It is important to evaluate the impact of changes to surgical training on patient outcome. Due to the highly technical nature of the surgery, rectal cancer surgery lends itself well to quality outcome assessment with respect to surgical training. Technical factors such as circumferential margin status [13] and quality of mesorectal dissection plane [14] have been shown to be associated with local recurrence. This study, therefore, sought to evaluate the outcome of elective primary excisional colorectal cancer surgery in patients treated by CTCs compared to NCTCs using a prospectively collected local clinicopathological dataset and nationally using an administrative dataset. This would allow us to investigate the more technical and oncological factors integral to rectal cancer surgery, which are omitted from the administrative nature of the Hospital Episode Statistics (HES) dataset.
Methods Training scheme definitions Historically, surgical training schemes in the UK adhered to the apprenticeship model, with no fixed training time, but were estimated to cover 12,000 h [2–7]. In 1996, the Calman reforms overhauled Specialist Registrar training with definition of minimal training scheme lengths and formal educational objective setting, training agreements and regional rotational placements [3]. Significant reductions in training hours of UK surgical trainees were documented with Calman trainees, down to approximately 8,000 h [8]. Local data protocol Between April 2006 and December 2010, all patients with operable rectal cancer (defined as within 15 cm of the anal verge) managed in a single NHS trust (two surgical units) were included. A single MDT managed all patients. Patients with involved or threatened CRMs as predicted by MRI staging were offered neoadjuvant chemoradiotherapy, as were all distal rectal cancers ≥T2 (less than 6 cm from anal verge) due to increased recurrence rates reported in the literature in this latter group of patients. Radiotherapy consisted in most cases of 45Gy in 28 fractions over 5 weeks with capecitabine Monday to Friday at 825 mg/m2. All other staged rectal cancers tumours were offered primary surgery unless they were entered into the CR07 trial [15]. A small number of cases received short-course RT outwith trials. Surgery was undertaken within 10 days of short-course or 6–8 weeks of long-course RT. Clinical, surgical and histopathological data were recorded prospectively in a registry database. Data fields included
Int J Colorectal Dis (2014) 29:631–638
gender, age, referral pathway, ASA grade, tumour localization, type of resection and operative approach (conversioncoded), intra- and post-operative complications (with return to theatre-coded as an intra-abdominal procedure or wound complication post-index procedure), anastomotic leak (defined as suspicious clinical examination findings and subsequently proven at CT imaging or laparotomy), number of resected lymph nodes, circumferential margin status, pathological TNM stage (version 5), tumour recurrence (local and systemic) and survival. Mortality data were gathered from the local cancer registry (Thames Cancer Registry). HES data The HES database has been described previously [15]. In brief, it is a record-based system that since 1986, has collected patient-level data from all English NHS Trusts. Each HES record contains geographic, demographic, diagnostic and procedural data pertaining to an individual patient’s admission to an NHS hospital. Each episode contains a primary and up to 19 secondary diagnoses, categorised according to the International Classification of Disease (10th revision) (ICD10), and up to 24 procedure fields coded using the Office of Population Censuses and Surveys Classification of Surgical Operations and Procedures (4th revision) (OPCS-4). The Charlson co-morbidity index, derived from the secondary diagnosis scores, was used as a measure of co-morbidity [16]. The Carstairs index of deprivation is a small area deprivation score that is based on the patient’s postcode [17]. Database inclusions and variable recoding All primary elective colorectal cancer resections carried out in English NHS Trusts between 1 April 2000 and 31 March 2008 were identified. As Calman training for surgical Specialist Registrars commenced nationally in 1997 and presuming a 7-year higher surgical training programme, any surgeon appearing on the HES dataset in 2004–2008 only was labelled as a CTC. Any surgeon operating within the period 2000–2004 was considered to be a NCTC. Patient outcomes were identified for those admitted between 1 April 2004 and 31 March 2008. To stratify patients, they were grouped by age into four clinically relevant categories: younger than 55 years, 55 to 69 years, 70 to 79 years and older than 79 years. Charlson index was grouped into three categories: zero (no comorbidity), one to four (moderate co-morbidity) and greater than or equal to five (severe co-morbidity). Outcome variables Thrity-day mortality was defined as death in hospital within 30 days of admission. Readmission was considered as an unplanned readmission within 28 days of discharge. Length of stay is the time (in days) that a patient remained in the
Int J Colorectal Dis (2014) 29:631–638
hospital during the index admission. Owing to its non-normal distribution, length of stay was analysed by log-normal transformation and independent t tests with back exponentiation to provide relative risk values. Reoperation was defined as any return to theatre for an intra-abdominal procedure or wound complication on the index admission or on a subsequent admission to hospital within 28 days of the initial resection [18]. Statistical methodology Categorical variables were investigated using the chisquare test. t tests and Kruskall Wallis were used to evaluate differences between continuous normally and non-normally distributed variables, respectively. Median (interquartile range) and mean (standard deviation) are described as appropriate. Logistic regression models were constructed to evaluate the independent role of Calman training, operative technique (i.e. laparoscopic versus open), operation type, age, gender, co-morbidity, year of surgery and social deprivation on 30-day mortality, readmission and reintervention rates. A multiple linear regression model was constructed to evaluate the role of Calman training, operative technique, operation type, age, gender, co-morbidity, year of surgery and social deprivation on the natural logarithm of length of stay. All covariates significant at p79 Resection type Right sided Left sided Rectal Subtotal/total Laparoscopic procedure Carstairs deprivation 1 (least deprived) 2 3 4 5 (most deprived) Charlson score 0 1–4 ≥5 30-day mortality Readmission Reintervention LOS (days); mean (SD)
9,594
34,512
891 (9.3) 3,405 (35.5) 3,459 (36.1) 1,839 (19.2)
3,298 (9.6) 12,155 (35.2) 12,434 (36.0) 6,625 (19.2)
0.87
3,335 (34.8) 1,709 (17.8)
10,354 (30.0) 6,874 (19.9)
ORIGINAL ARTICLE
The impact of shortened postgraduate surgical training on colorectal cancer outcome A. Currie & E. M. Burns & P. Aylin & A. Darzi & O. D. Faiz & P. Ziprin & on behalf of the Imperial College Healthcare NHS trust Colorectal Cancer MDT
Accepted: 11 February 2014 / Published online: 6 March 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Background Shortened postgraduate surgical training reforms, known as Calman, have altered delivery of surgical training in the UK with reduced working hours and training time aiming to produce a more subspecialised workforce. Aims This study aims to compare rectal cancer surgical outcomes of Calman-trained consultants in a single institution to published data. Additionally, the study compared colorectal cancer surgical outcome between Calman-trained consultants (CTCs) and non-Calman consultants (NCTCs) in a national dataset. Methods Local dataset Clinicopathological outcome of rectal cancer resection undertaken by CTCs in a single institution (2006–2010) were compared against NCTC counterparts. National dataset All elective colorectal cancer resections between 2004 and 2008 in English NHS hospitals were included. CTCs (present from 2004 onwards) were compared to NCTCs (present prior to 2004). Outcome measures included 30-day in-hospital mortality, reoperation and readmission rates. Results Local dataset One hundred thirteen patients were operated under five CTC. The 30-day in-hospital mortality Presented in poster form at the Association of Surgeons of Great Britain and Ireland International Surgical Congress, Bournemouth, UK. May 2011 What’s new in this paper? This paper is the first to examine the outcomes of shortened postgraduate training methods in outcomes in the setting of predominantly surgically managed rectal cancer. A. Currie : E. M. Burns : A. Darzi : O. D. Faiz : P. Ziprin (*) Department of Surgery, Imperial College, St Mary’s Hospital, Praed Street, London W21NY, UK e-mail: [email protected] P. Aylin Dr. Foster Unit, Department of Primary Care and Social Medicine, Imperial College, 1st Floor Jarvis House, 12 Smithfield Street, London EC1A 9LA, UK
for CTCs (1 %) was favourable compared to published rates (3–5 %). Local recurrence rate (4.4 %) was comparable to NCTC (3.6 %). National dataset Between 2004 and 2008, 44,106 patients underwent elective colorectal resection. Multiple regression demonstrated CTC patients had a reduced length of stay and reduced reoperation rate. No difference in mortality and unplanned readmission rates were seen. Conclusion CTCs have similar safety outcome to NCTCs for colorectal cancer resection procedures. Further work is needed to assess the impact of further training reductions on clinical outcome. Keywords Calman . Colorectal cancer . Postgraduate surgical training
Introduction Quality assurance has become an increasingly important aspect of colorectal surgery. Technical outcomes, such as circumferential resection margin (CRM) status [1], have been suggested as potential quality measures. Significant changes have occurred to the technical training of surgeons in the UK [2–8], Australasia [9, 10] and North America [11, 12] and led to a reduction in training hours. One of the aims of such reforms is to deliver a more specialised workforce. The Calman reforms, fully instituted in 1997, overhauled UK postgraduate medical training [3] in an attempt to deliver a more subspecialised surgeon. Similar reforms took place in the European and US training systems. These changes resulted in concerns regarding reduced surgical trainee operating volume [8], shorter overall surgical training and the potential impact on patient outcome [2–7]. To date, no studies have compared outcome amongst Calman-trained consultant
632
(CTC) surgeons with non-Calman consultant (NCTC) counterparts. It is important to evaluate the impact of changes to surgical training on patient outcome. Due to the highly technical nature of the surgery, rectal cancer surgery lends itself well to quality outcome assessment with respect to surgical training. Technical factors such as circumferential margin status [13] and quality of mesorectal dissection plane [14] have been shown to be associated with local recurrence. This study, therefore, sought to evaluate the outcome of elective primary excisional colorectal cancer surgery in patients treated by CTCs compared to NCTCs using a prospectively collected local clinicopathological dataset and nationally using an administrative dataset. This would allow us to investigate the more technical and oncological factors integral to rectal cancer surgery, which are omitted from the administrative nature of the Hospital Episode Statistics (HES) dataset.
Methods Training scheme definitions Historically, surgical training schemes in the UK adhered to the apprenticeship model, with no fixed training time, but were estimated to cover 12,000 h [2–7]. In 1996, the Calman reforms overhauled Specialist Registrar training with definition of minimal training scheme lengths and formal educational objective setting, training agreements and regional rotational placements [3]. Significant reductions in training hours of UK surgical trainees were documented with Calman trainees, down to approximately 8,000 h [8]. Local data protocol Between April 2006 and December 2010, all patients with operable rectal cancer (defined as within 15 cm of the anal verge) managed in a single NHS trust (two surgical units) were included. A single MDT managed all patients. Patients with involved or threatened CRMs as predicted by MRI staging were offered neoadjuvant chemoradiotherapy, as were all distal rectal cancers ≥T2 (less than 6 cm from anal verge) due to increased recurrence rates reported in the literature in this latter group of patients. Radiotherapy consisted in most cases of 45Gy in 28 fractions over 5 weeks with capecitabine Monday to Friday at 825 mg/m2. All other staged rectal cancers tumours were offered primary surgery unless they were entered into the CR07 trial [15]. A small number of cases received short-course RT outwith trials. Surgery was undertaken within 10 days of short-course or 6–8 weeks of long-course RT. Clinical, surgical and histopathological data were recorded prospectively in a registry database. Data fields included
Int J Colorectal Dis (2014) 29:631–638
gender, age, referral pathway, ASA grade, tumour localization, type of resection and operative approach (conversioncoded), intra- and post-operative complications (with return to theatre-coded as an intra-abdominal procedure or wound complication post-index procedure), anastomotic leak (defined as suspicious clinical examination findings and subsequently proven at CT imaging or laparotomy), number of resected lymph nodes, circumferential margin status, pathological TNM stage (version 5), tumour recurrence (local and systemic) and survival. Mortality data were gathered from the local cancer registry (Thames Cancer Registry). HES data The HES database has been described previously [15]. In brief, it is a record-based system that since 1986, has collected patient-level data from all English NHS Trusts. Each HES record contains geographic, demographic, diagnostic and procedural data pertaining to an individual patient’s admission to an NHS hospital. Each episode contains a primary and up to 19 secondary diagnoses, categorised according to the International Classification of Disease (10th revision) (ICD10), and up to 24 procedure fields coded using the Office of Population Censuses and Surveys Classification of Surgical Operations and Procedures (4th revision) (OPCS-4). The Charlson co-morbidity index, derived from the secondary diagnosis scores, was used as a measure of co-morbidity [16]. The Carstairs index of deprivation is a small area deprivation score that is based on the patient’s postcode [17]. Database inclusions and variable recoding All primary elective colorectal cancer resections carried out in English NHS Trusts between 1 April 2000 and 31 March 2008 were identified. As Calman training for surgical Specialist Registrars commenced nationally in 1997 and presuming a 7-year higher surgical training programme, any surgeon appearing on the HES dataset in 2004–2008 only was labelled as a CTC. Any surgeon operating within the period 2000–2004 was considered to be a NCTC. Patient outcomes were identified for those admitted between 1 April 2004 and 31 March 2008. To stratify patients, they were grouped by age into four clinically relevant categories: younger than 55 years, 55 to 69 years, 70 to 79 years and older than 79 years. Charlson index was grouped into three categories: zero (no comorbidity), one to four (moderate co-morbidity) and greater than or equal to five (severe co-morbidity). Outcome variables Thrity-day mortality was defined as death in hospital within 30 days of admission. Readmission was considered as an unplanned readmission within 28 days of discharge. Length of stay is the time (in days) that a patient remained in the
Int J Colorectal Dis (2014) 29:631–638
hospital during the index admission. Owing to its non-normal distribution, length of stay was analysed by log-normal transformation and independent t tests with back exponentiation to provide relative risk values. Reoperation was defined as any return to theatre for an intra-abdominal procedure or wound complication on the index admission or on a subsequent admission to hospital within 28 days of the initial resection [18]. Statistical methodology Categorical variables were investigated using the chisquare test. t tests and Kruskall Wallis were used to evaluate differences between continuous normally and non-normally distributed variables, respectively. Median (interquartile range) and mean (standard deviation) are described as appropriate. Logistic regression models were constructed to evaluate the independent role of Calman training, operative technique (i.e. laparoscopic versus open), operation type, age, gender, co-morbidity, year of surgery and social deprivation on 30-day mortality, readmission and reintervention rates. A multiple linear regression model was constructed to evaluate the role of Calman training, operative technique, operation type, age, gender, co-morbidity, year of surgery and social deprivation on the natural logarithm of length of stay. All covariates significant at p79 Resection type Right sided Left sided Rectal Subtotal/total Laparoscopic procedure Carstairs deprivation 1 (least deprived) 2 3 4 5 (most deprived) Charlson score 0 1–4 ≥5 30-day mortality Readmission Reintervention LOS (days); mean (SD)
9,594
34,512
891 (9.3) 3,405 (35.5) 3,459 (36.1) 1,839 (19.2)
3,298 (9.6) 12,155 (35.2) 12,434 (36.0) 6,625 (19.2)
0.87
3,335 (34.8) 1,709 (17.8)
10,354 (30.0) 6,874 (19.9)
