The main guideline developer group consisted of 14 people, including physicians, physiatrists, orthopedists, vascular surgeons, experts in prosthetics, psychiatrists, physiotherapists, occupational therapists, clinical epidemiologists, public health doctors, economists, a documentary librarian, and undergraduate and postgraduate students. A group of professionals from different universities and scientific societies validated the different stages of the process. A focus group of 24 people with amputations of different causes and their relatives were linked to the process in two moments, when the questions were chosen and at the end of the recommendations. The developer group received a training process with different international centers as the McMaster University, the National Institute for Health and Clinical Excellence (NICE) and the New Zealand Guidelines Group. The users of the CPG are all the professionals who were involved in the development: surgeons, physiatrists, other professionals in the area of rehabilitation, insurers, health providers and political decision makers.

All the professionals who participated in the development made a declaration of interests at the beginning and each year. These are published as supplementary files within the CPG document (8).

The financing entity was the Ministry of Health and Social Protection, none of the people from this entity participated in the group developing the CPG. The Ministry carried out permanent monitoring to guarantee compliance with the methodology and schedules.

CPG for lower limb amputees were searched for, and an evaluation of quality was made with the AGREE II Instrument. Six CPG were evaluated independently by two professionals from the group. Only three with a score greater than 60 in the methodological domain were selected, which were used as information during development. (8).

The development of the CPG followed the GRADE (Grading of Recommendations Assessment, Development and Evaluation) methodology (10, 11). For the elaboration of the recommendations of the CPG, within the guideline development group (GDG), a process of prioritization of the topics of interest was carried out and the most important were selected. Subsequently, clinical questions were formulated and a systematic review of the available evidence was made on each one. This process was done between 2014 and 2015. The main recommendations were updated in 2018.

Based on the scope, purposes and objectives of the guideline, the questions were elaborated with the PECOT strategy (Population, Exposure or intervention, Comparison, Outcomes and Time). Then the developer group and the patients independently rated the importance of each outcome from 1 to 9, according to the GRADE classification (11). According to the average scores of the developer group, the outcomes were classified as: critical (7-9), important non-critical (4-6) or not important (1-3). The evaluation of the quality of the body of evidence is done by selecting the critical and important outcomes.

For each question, a list of MeSH terms was prepared according to the population, the intervention and the comparison. The evidence search was performed in the main databases: Cochrane Library, Embase and PubMed, and in secondary databases such as Lilacs/Bireme, Current Controlled Trials, TripDatabase and Google Scholar. There was no language restriction. For the selection of the evidence, inclusion criteria were established with respect to the methodological design, the population and the minimum quality characteristics. Systematic reviews and meta-analyses (secondary or aggregative studies), which analyzed primary studies related to the question, were initially sought. Additionally, clinical trials and observational studies were identified.

The quality of the evidence was evaluated for systematic reviews and meta-analyses with the AMSTAR (12); for diagnostic studies the QUADAS (13); and with the STROBE for observational studies (14). The quality of the body of evidence was assessed according to the concepts of the GRADE methodology (11), by qualifying each outcome. This process was done by orthopedic doctors, physiatrists, and clinical epidemiologists, who were experts in the GRADE methodology. GRADE publications can be accessed on the website https://www.gradeworkinggroup.org/.

Systematic reviews of clinical trials start with high quality (level 1), while reviews of observational studies start with low quality (level 4). The aspects that can lower the quality of a randomized controlled trial are: Risk of bias, inconsistency of the results, indirect evidence, imprecision of the results and publication bias. Observational studies, although they can lower quality with the aforementioned aspects, also they can increase it if they include some favorable methodological aspects. The three aspects that can increase the quality of are the presence of a large effect size (Relative Risk, >2.0 or <0.5); evidence of a gradient dose-response relationship and the absence of residual bias or confounding factors (15–20).

The quality of the evidence is related to the confidence that the true effect is close to the estimated effect. Four levels are defined: very low, low, moderate and high (11). Most of the quality of the evidence for this guideline was low or very low quality of evidence.

Following the GRADE system, the elaboration of the recommendations does not only take into account the quality of the evidence, but also a series of aspects or judgments based on the following criteria: The priority of the problem, the magnitude of the desirable and undesirable effects, the certainty of the evidence, the values of the interested parties, the balance between desirable and undesirable effects, the resources required, the cost-effectiveness, equity, acceptability and feasibility. With these criteria, a summary table of judgments was created and the direction and strength of the recommendation were defined (21). The strength of the recommendations is rated in four categories: Strong (recommended to do), weak in favor (suggested to do), strong against (recommended not to do), weak against (suggested not to do) (21). During face-to-face sessions with the entire guideline development group, the evidence for each question, the quality of the body of evidence, and the judgments were presented. With the foregoing, a recommendation was drawn up that was subsequently validated by an extended group with thematic experts and representatives of scientific societies and universities.

Five economic evaluations were made during the development of the CPG to assess the cost-effectiveness of five of the interventions and help the guideline development group in the decision-making.

In the development of a CPG, it is recommended including the perspective of the patients for the preparation of the recommendations. Thus, people with lower limb amputation were invited to define their priorities in three categories: complications, activities and prosthetic adaptation; using the GRADE methodology. In addition, their preferences of the treatment options in the recommendations with greater uncertainty and with low quality of evidence were evaluated.

Between July and November 2014, people with amputation in two institutions that provide health services in two cities of the country were invited. The inclusion criteria were people from 18 to 65 years old, who had a major lower limb amputation of any level and cause and who could attend a meeting with the researchers. Children and upper limb amputees were excluded. A convenience sampling was used, with the people who responded to the call. The objectives of the CPG and their participation, the instruments that were applied and doubts were resolved were explained at the meeting. In the group of 20 patients studied in one of the cities, the preferences of the CPG questions in which there was greater uncertainty at the time of presentation of the evidence synthesis were also evaluated.

One search strategy is presented as an example in Table 1, for one of the surgical recommendations, elaborated during the updating of the CPG. All the other search strategies can be consulted in the complete document of the CPG (8).

Forty-three recommendations were made. Nine on the decision of amputation; five on preoperative interventions including: preoperative regional analgesia, cardiovascular reconditioning, psychological support, prophylactic antibiotics, and intraoperative tourniquet use; ten on amputation techniques; ten on the components of the prosthesis, feet, knees, sockets, liners, as well as orthoses for partial amputation and the adaptation of immediate postoperative prostheses. Nine for the post-prosthetic phase, including functioning scales to evaluate the use of prostheses; treatment for neuropathic and phantom limb pain; cardiopulmonary, physical and occupational rehabilitation; ergonomic adaptations and psychosocial interventions. Comprehensive rehabilitation compared to the usual care model was also evaluated and this was a strong recommendation in favor. The synthesis of the quality of the evidence and the strength of the recommendations can be seen in Table 2, and Figures 1, 2.

The distribution of the 43 recommendations, according to the quality of the evidence, was: high 0%, moderate 7.5%, low 57.5%, very low 35%. And according to the strength of the recommendations, their distribution was: Strongly in favor 37.2%; weak in favor 53.48%; strong against 2.32%; weak against 6.97%. In updating the prioritized questions, a recommendation changed from Weak in favor to Strong in favor, leaving the distribution as follows: Strong in favor 58.3%, weak in favor 41.6%.

As part of the participation of patients in the development of the CPG, they were invited to assess the importance of each outcome of the recommendations (22). Patients chose stump infection in 31.7%, death in 22%, stump reoperation in 22%, and phantom pain in 12.2% as the most important outcomes for them. The most important activities in the rehabilitation process were walking in 51.2%, returning to work in 17.1%, having a good quality of life in 14.6% and participating in social activities in 7.3%. Twenty patients participated in the evaluation of preferences. Of them, 95% prefer to keep the knee to a transfemoral amputation, 60% prefer amputation in the first surgery than reconstruction, 75% agree with the need for psychological support, 85% agree with a supervised exercise plan after amputation and, only 45% agree with the use of an immediate prosthesis (22).

The results of the first economic evaluation concluded that, after a careful selection of patients and intervention by a multidisciplinary team, limb reconstruction was a dominant strategy compared to primary amputation in the long term (23).

In the second evaluation, the adaptation of an articulated foot was not a cost-effective strategy compared to a SACH foot, in patients with a low level of activity (8).

In the third, in a sample of 113 patients analyzed in a cross-sectional study, the total contact socket was a cost-effective strategy compared to a patellar-tendon-bearing (PTB) type (8). However, it is not possible to determine if this result can be extrapolated to other populations of patients with amputations below the knee in Colombia.

In the fourth economic evaluation, Pregabalin was found to be the strategy with the greatest net benefit, so it can be considered first-line treatment of phantom pain or residual pain in lower limb patients with amputation (24). Gabapentin and amitriptyline had similar, albeit lower, net benefits and could also be considered at the discretion and experience of the treating physician. More research is needed on the effectiveness of medications in patients with lower limb amputation.

In the fifth economic evaluation on the cost-effectiveness of prophylactic antibiotics, it was found that this is a dominant strategy and it is unlikely that the uncertainty surrounding the costs and benefits changes the results, the use of this intervention is recommended in Colombia (25).

As a final result of the CPG, the implementation process was described based on planning, implementation activities, monitoring and evaluation. Structure, process and outcome indicators were defined. Structure indicators were the availability of surgical and rehabilitation services. Process indicators were the proportion of patients with prostheses and in a rehabilitation program, according to the recommendations of the guidelines. Outcome indicators were the proportion of patients with reamputation, the proportion of patients adapted to prosthesis, and the proportion of professionals who follow the recommendations of the CPG.

For health professionals in charge of caring for amputees, it is important to emphasize the need for patients to receive the most effective and safe interventions. Patients need to receive this intervention in time to reduce functional limitations and achieve occupational reintegration and social participation for amputees.

Rehabilitation services must be comprehensive and available in a place that is close to patients to reduce the possibility of loss in the continuity of care. In the country, travel is paid for by patients and their families and this can be an even greater barrier if they must go to different places for their treatment. Comprehensive rehabilitation must involve mental health aspects to prepare amputees in the phase of acceptance and mourning for the loss of their limb and provide support in rehabilitation. Also, comprehensive rehabilitation must include physiatrist care, physical and occupational therapy, cardiopulmonary rehabilitation, psychology care and very importantly access to the prosthesis and its adaptation. All of the above will make possible for amputees to return to their usual occupation and integrate into society.

Professionals must have the necessary training, time and incentives to achieve a change in professional practice.

In Colombia, administrative procedures with insurers companies are lengthy and amputee patients must take multiple steps to obtain approval for each of the interventions and devices necessary for their rehabilitation (36, 37). The rehabilitation program should be approved as a package of interventions based on the recommendations of the CPG.

One limitation of the study is not having the final results of the implementation project to make better analyzes of the situations presented.