Introduction

Colorectal cancer (CRC) is a very common cancer, with 1.8 million new cases registered worldwide in 2018 [1]. The preferred curative treatment option is surgical resection, when indicated in concurrence with (neo)adjuvant therapy. Timely diagnosis and start of treatment have become important goals in optimizing outcomes. The interval between diagnosis and treatment is subject of debate since a prolonged interval may negatively affect oncological outcome. Recommendations regarding length of the CRC treatment interval are incorporated in guidelines in various countries. Those recommendations are often used as an indicator for quality of care and as a surrogate measure of the effectiveness of cancer services [2, 3]. Even so, not meeting the recommendation might result in consequences. For example, in the UK a financial penalty can be imposed by the Clinical Commissioning Group [4, 5]. However, the rationale for these national guidelines is mainly based on consensus and expert opinion only [6, 7].

It is widely accepted that prehabilitation enhances functional capacity prior to CRC treatment and improves postoperative outcome [8,9,10,11,12]. The interval between diagnosis and treatment could thus be used to implement a multimodal prehabilitation program. However, the recommendations for length of the treatment interval may hinder professionals to implement such a program.

In order to find out the definition of the interval upon treatment—“the treatment interval”—and whether this treatment interval could be safely used to implement prehabilitation in CRC, we addressed the following questions:

  1. (1)

    What is the definition of the CRC treatment interval?

  2. (2)

    What are the recommendations for CRC treatment interval length included in national guidelines worldwide and are those recommendations feasible?

  3. (3)

    What is the possible association between outcome and length of the interval between diagnosis and treatment?

Material and methods

Data were collected through a literature review, a web search and a survey among colorectal surgeons worldwide. The literature review provided results for all three research questions. The web search and survey provided additional information about the CRC treatment interval recommendations included in international guidelines discussing timelines (research question 2).

Literature review

A systematic search was performed of the following electronic databases: MEDLINE (1946 to 2020 June 3), EMBASE (1974 to 2020 June 3) and the Cochrane Library (1992 to 2020 June 4) including the following search terms “colorectal neoplasms,” “time-to-treatment,” “time factors” and “waiting lists” (complete search string displayed in Online Resource 1). Titles and abstracts of all records identified by the search were independently screened and assessed for eligibility by two authors (CM and LJ). Articles were deemed eligible if they (1) described CRC either specifically or in combination with other diseases and (2) included a description of the treatment interval defined as any time point in the cancer care pathway until initiation of any form of CRC treatment. The search was restricted to English and Dutch written papers with no limitation in date or study design. Papers were excluded when meeting the following exclusion criteria: interval described with an ending other than treatment, interval described between treatment modalities (e.g., time between neoadjuvant treatment and surgery or between surgery and adjuvant therapy). Full-text articles were retrieved when a paper was considered eligible based on title and abstract or when information was insufficient to determine eligibility. Additionally, bibliographies of included studies were hand-searched to identify any further eligible studies. Any disagreements were discussed. When discordance continued, a third author (GS) arbitrated until consensus was reached. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline was used as guidance for reporting the current systematic review [13].

The following data were extracted independently by CM and LJ using a predefined collection form:

  • General information: first author, publication date, country, journal;

  • Study characteristics: disease(s), study design, sample size, recommendation on length of treatment interval described in the paper, outcomes and results.

Furthermore, specific information was extracted for each research question. For research question 1 regarding the definition of the treatment interval:

  • What time points of the CRC care pathway were used to define the start and end of the treatment interval? For the scope of this question, treatment interval was considered as the time period between any time point in the cancer care pathway until initiation of any form of treatment;

  • Secondly, how were the mentioned time points determined;

  • For data extraction: when two definitions of the treatment interval and/or time point were mentioned in a paper, both definitions were registered. However, in case the paper referred to a hierarchical definition of a time point (e.g., the European Network of Cancer Registries hierarchy [14]), only the upper preferred definition was recorded;

  • A definition of the treatment interval was deemed complete when the following information was provided in the paper: (1) a description of what time points in the cancer care pathway were used to define the start and end of the interval and (2) a description of how these specific time points were determined.

For research question 2 regarding the recommendations on length of the treatment interval included in guidelines:

  • Guideline recommendations and success rate (percentage the aimed target recommended in the guideline was met) described in the paper were only registered when this guideline was actually effectuated in the country the paper originated from.

For research question 3 regarding the association between outcome and length of the interval between diagnosis and treatment:

  • Only the papers using the time point “diagnosis” as start of the treatment interval were included for this question;

  • In case length of the treatment interval was reported for several treatment types, time to surgery was chosen. When length of treatment interval was reported over a period of time, data of the most recent year were extracted. If possible, length of treatment interval without urgent treatments was reported. Finally, when length of treatment interval was described for a standard and an interventional pathway (e.g., direct access colonoscopy versus standard referral), results for the standard pathway were extracted.

Web search

A search on the World Wide Web was performed to complement the overview of international guidelines containing recommendations on length of the CRC treatment interval. The following terms were used to search the web: “colorectal” or “bowel” and “cancer” or “carcinoma” combined with “treatment interval” or “waiting time” and “target,” “recommendation” or “guideline.” Websites of Ministries of Health, cancer societies and colleges of specialists were screened. The web search was restricted to websites written in English or Dutch.

International survey

Since limited countries were represented in the literature and on the web search, additional information on the guidelines in various countries was retrieved by conducting a survey among colorectal surgeons from countries worldwide.

The authors designed a survey based on collected information (Online Resource 2). Intervals, time points and definitions described in the literature or online were used to provide various options regarding the treatment interval. It was arbitrarily decided to use our own network consisting of experts in the field in 33 countries. Based on our information, these surgeons are currently regarded as experts in the field of colorectal oncology by both their peers as well as by the international community. All of them have published studies in international literature and are currently engaged in the treatment of these populations. Moreover, they were willing to act as a representative for their country and provide the required information within a short period of time. Conversely, by randomly approaching national committees or medical societies, we were not convinced that we would receive the proper information in due time. The survey was sent by email, and in case of nonresponse, a reminder was sent after two weeks.

Results

The search in the electronic databases MEDLINE, EMBASE and Cochrane Library on June 3–4, 2020 together with the search of bibliographies resulted in 110 included papers. For the first research question, 106 papers were included, 39 papers for question 2 and 30 papers for question 3. There is overlap in those numbers since some papers contained data for more than one research question. The screening and selection process is displayed in a PRISMA flow diagram (Fig. 1) [15].

  1. 1.

    What is the definition of the CRC treatment interval?

Fig. 1
figure 1

PRISMA flow diagram

Full size image

Of the included articles, 106 contained a description of the treatment interval. Five papers included two separate definitions of the treatment interval resulting in 111 intervals described. Time points in the cancer care pathway used to define start and end of the interval were mentioned in all 111 intervals (Table 1). For 63 intervals, a complete definition was provided, containing both definition and determination of time points. The definition and determination of the time points varied widely. The treatment interval generally started with “diagnosis” and often ended with “treatment in general.” Subsequently, date of “diagnosis” was generally determined using date of clinical and/or pathological confirmation and for date of “treatment in general” date of initiation of any treatment was used in the majority of the papers.

  1. 2.

    What are the recommendations for CRC treatment interval length included in national guidelines worldwide and are those recommendations feasible?

Table 1 Definition and determination of the CRC treatment interval and associated time points of the cancer care pathway in 106 included papers
Full size table

Thirty-nine included papers from the literature review described treatment interval recommendations or the lack thereof. Papers also reported the success rate, i.e., percentage the aimed target recommended in the guideline was met.

The survey on treatment interval recommendations was sent to 33 surgeons in 33 countries, and 22 surgeons completed the survey. Data from the survey are only displayed when recommendations differed from the results found through the literature review or web search, or when recommendations for the specific country were unknown.

Guidelines differed between countries regarding the definition of the treatment interval (Table 2). Additionally, the recommendations on the timeline of this interval varied as well, ranging from surgery within two weeks from decision to treat, to treatment within seven weeks from pathological confirmation. Some countries have more than one guideline containing different recommendations on length of the treatment interval, others have no guideline (yet). The majority of the recommendations were based on expert opinion. The recommended targets of these guidelines were met in 21–80%.

  1. 3.

    What is the possible association between outcome and length of the interval between diagnosis and treatment?

Table 2 Recommendations for treatment interval included in (inter)national guidelines, displayed per country and reported percentages those recommendations were met
Full size table

In 30 papers, the association between length of the treatment interval (diagnosis—treatment) and outcome in colon, rectal or colorectal cancer was studied (Table 3). The majority of included studies (n = 19) did not find an association between length of CRC treatment interval and outcome. Six papers concluded that a long treatment interval for colonic and/ or rectal cancer treatment is associated with worse outcome than a short treatment interval. Length of the treatment interval described as “long” in those papers ranged from > 31 to > 84 days. Meanwhile, one paper concluded that a long CRC treatment interval is associated with better outcome. Finally, four papers described a U-shaped association with worse outcome when length of the treatment interval was either short or long, compared to an intermediate length. Outcome included surgical outcome (e.g., length of hospital stay, complication rate) or oncological outcome (e.g., tumor stage, recurrence rate, or survival). Survival included 5-year survival, overall survival and disease-specific survival.

Table 3 Papers studying the association between length of the treatment interval and outcome in CRC
Full size table

Discussion

Based on this study, we conclude that a uniform description of the treatment interval for CRC is absent. Time points defining the start of the interval are not consistently determined. Guidelines and their recommendations for this interval differ widely among countries. Finally, there is no evidence that a short interval between diagnosis and treatment improves oncological outcome.

Heterogeneity in the reporting of cancer care pathway intervals has also been described by other authors [80,81,82,83]. We found that the end of the treatment interval is rather consistent defined as date of initiation of treatment. However, variety in the start of the treatment interval is high ranging from onset of symptoms to definitive diagnosis. Date of diagnosis is often used as start of the interval; however, a clear description of what event represents “diagnosis” is then often unclear and even not specified in nearly one-third of the papers. A clear definition is a necessity since, for example, date of diagnosis and consequently the treatment interval might differ a full week when date of biopsy or date of final pathology report is used [80, 82]. This makes comparison of this quality indicator among institutes difficult. Date of pathological confirmation is a clear and internationally applicable time point in the cancer care pathway and is often used in the literature as date of diagnosis. We therefore recommend the following universal definition for the CRC treatment interval: the interval starts with date of diagnosis, determined by date of pathological confirmation, and the interval ends with the date any primary treatment for CRC is initiated.

Timely diagnosis and treatment is thought to improve oncological outcome. Based on the results of this study, we conclude that length of the CRC treatment interval—starting with date of diagnosis—is not associated with worse outcome. The reported association between a short treatment interval and worse outcome is likely caused by selection; patients with poorer clinical condition or a complication of CRC are generally treated in an emergency setting. Of the papers describing worse outcome with a long treatment interval (including a U-shaped association), only three papers relate a relatively short treatment interval of approximately four weeks to worse outcome [32, 76, 78]. The remaining papers described similar findings with longer intervals ranging from five to 13 weeks. An interval of five weeks from diagnosis to treatment seems safe. Hangaard Hansen et al. even suggested that an interval of eight to nine weeks between diagnosis and treatment is safe regarding the long-term oncological outcomes in colonic cancer [7]. Despite these results, a prolonged treatment interval may worry patients. Healthcare professionals should therefore inform patients optimally regarding the expectations of timeliness in perspective of its meaning to the disease and process [84].

Since timeliness of treatment has become a fundamental objective to ensure quality of care, guidelines including recommendations have been published worldwide. The majority of the recommendations for the treatment interval are based on consensus and expert opinion [6, 7] and are drafted by Ministries of Health, cancer or Medical Societies. Recommendations take the possible distress of the patient, caused by a long interval, into account. However, beside timeliness [84], interpersonal skills of the treating physician and coordination of care have a large influence on the patients’ satisfaction with waiting times [85]. This study found that a universal recommendation for the treatment interval is not available. Guidelines differ between countries, and some countries remarkably have more than one guideline, containing distinct recommendations. The recommended time to treatment ranges from two to seven weeks from decision to treat and pathological confirmation, respectively.

Ideally, the recommended lengths of the treatment interval are feasible in practice, especially considering the potential consequences when timelines do not comply with those recommendations [4, 5]. However, the reported success rates of 21–80% suggest that guidelines are not in line with practice. When these recommendations are considered indispensable, at least a uniform and feasible guideline should exist, leaving room for the professionals to deviate from this guideline. To our knowledge, there is no global initiative to draft such a uniform guideline.

Several limitations apply to the current study. Since CRC and waiting time were the focus in the search string, papers containing information on the CRC treatment interval may be missed when this was not the main subject of the paper or were included in the key words but rather mentioned as a detail. However, due to the systematic conduct including a thorough search of the bibliographies, the amount of missed records is expected to be minimal. Preferably, only papers reporting on the interval to primary treatment for elective CRC cases were included. Since the reporting on this subject is heterogenic and sometimes incomplete, this was not possible. The systematic approach of data extraction and reporting as described in the methods section diminished variety in the results. Because of the language restriction applied to this study, we conducted a survey to collect information from other countries to complement the overview. Another limitation is the retrospective nature of most of included papers assessing the association between CRC treatment interval length and outcome. However, a randomized design could be deemed unethical. This is therefore the highest level of evidence available. Finally, due to heterogeneity of studies, stratification of patients by tumor stage or location was not possible, which assumably affects outcome.

The strength of the current paper is the complete evaluation of the CRC treatment interval. Previously published systematic reviews did study the association between length of the treatment interval, and outcome, however, was not corrected for heterogeneity in the use of recommendations and definitions regarding the start of the treatment interval. Furthermore, this paper displays the recommendations of nearly 30 countries worldwide.

We focussed on the treatment interval for CRC. CRC develops slowly over time and may take up to 10–15 years before it is diagnosed [86, 87]. Furthermore, 70% of total delay until treatment is determined by the period prior to diagnosis [28, 77, 88,89,90,91]. Based on these findings, one can assume that extending time to treatment (with a reasonable amount of time) will not harm the patient. Several authors state that time frames should be flexible in order to improve a patients’ functional capacity without detrimental effects on outcome [6, 18, 36, 71, 92]. Preoperative optimization can be achieved with a multimodal prehabilitation program to enable a patient to recover faster and better, with less complications and perhaps with an improved disease-free survival [11, 12]. In other words, delaying surgery when preoperative optimization is indicated rather than nationally applied treatment goals could benefit the patient [93].

Distinct from the previously suggested start of the treatment interval, namely pathological confirmation, prehabilitation could already start earlier in the cancer care pathway. Endoscopists are capable of determining (pre)malignant lesions that need a full work-up with approximately 90% accuracy [86]. This time point often initiates the oncological care pathway including work-up. Endoscopists should be able to initiate steps toward treatment and initiate prehabilitation after date of endoscopy. In order to maximize the time available for prehabilitation, work-up until diagnosis should be arranged effectively. The amount of time possibly gained by shortening the period for work-up can be used for prehabilitation without delaying time to treatment further. Some papers even suggest to consider prehabilitation as the start of initial treatment in CRC care and as an addition to anticancer therapy regimen [7, 10].

Finally, in the current study CRC is considered as a single tumor entity. Rectal cancer surgery is often preceded by neoadjuvant treatment, and surgery is generally more radical. The differences in cancer care pathways should be taken into account when implementing a prehabilitation program.

We conclude that there is no uniform definition of the CRC treatment interval. There is no decisive evidence for an association between length of the treatment interval and outcome in CRC. Justification to consider this as a quality measure and penalize institutes not meeting this criterion is therefore questionable. Furthermore, recommendations for CRC treatment interval length included in guidelines vary widely worldwide. Meanwhile, flexibility in the length of the CRC treatment interval enables professionals to implement prehabilitation in order to improve preoperative functional capacity and outcome.