The 2 most significant complications of colonoscopy with polypectomy are bleeding and perforation. Incidence rates for bleeding (0.1%–0.6%) and perforation (0.7%–0.9%) are generally low. Recognition of pertinent risk factors helps to prevent these complications, which can be grouped into patient-related, polyp-related, and technique/device-related factors. Endoscopists should be equipped to manage bleeding and perforation. Currently available devices and techniques are reviewed to achieve hemostasis and close colon perforations.
Key points
- •
The major complications of colonoscopy with polypectomy are postpolypectomy bleeding and colon perforation.
- •
Risk factors for postpolypectomy bleeding can be categorized as polyp-related, patient-related, and technique/device-related factors.
- •
Postpolypectomy bleeding can be controlled with injectable solutions and thermal and mechanical hemostatic devices.
- •
Risk factors for colon perforation can be polyp-related or technique/device-related.
- •
Methods of closure of colon perforation include endoclip placement, suturing, and others; appropriate management of patients with perforation is essential.
Introduction
Approximately 1.27 million colonoscopies are performed annually for colorectal cancer sreening. The risk of complications increases in patients who undergo colonoscopy with polypectomy ( Figs. 1 and 2 ). The rate of serious complications (ie, requiring hospitalization within 30 days of colonoscopy) is 10-fold higher (7.0 per 1000 examinations) if biopsy or polypectomy is performed compared with colonoscopy without these interventions. The 2 most significant complications include postpolypectomy bleeding (PPB) and colon perforation. The goal of this article is to review these complications in detail, assess risk factors for bleeding and perforation, and make recommendations regarding prevention and management.
Introduction
Approximately 1.27 million colonoscopies are performed annually for colorectal cancer sreening. The risk of complications increases in patients who undergo colonoscopy with polypectomy ( Figs. 1 and 2 ). The rate of serious complications (ie, requiring hospitalization within 30 days of colonoscopy) is 10-fold higher (7.0 per 1000 examinations) if biopsy or polypectomy is performed compared with colonoscopy without these interventions. The 2 most significant complications include postpolypectomy bleeding (PPB) and colon perforation. The goal of this article is to review these complications in detail, assess risk factors for bleeding and perforation, and make recommendations regarding prevention and management.
Postpolypectomy bleeding
The expected incidence of PPB ranges from 0.1% to 0.6%. A rate greater than 1% in a practice should prompt a review of endoscopist technique. PPB can be classified as immediate (intraprocedural or within 24 hours of the examination) or delayed (between 1 day to 14 days after the examination) and the severity of bleeding can be graded as follows: grade 1 (spontaneous hemostasis within 1 minute); grade 2 (continuous oozing but decreases over 1 minute); grade 3 (continuous oozing over 1 minute that requires endoscopic treatment); or grade 4 (active spurting) ( Box 1 ).
Grade 1: Spontaneous hemostasis within 1 minute of observation
Grade 2: Continuous oozing but decreases over 1 minute of observation
Grade 3: Continuous oozing over 1 minute of observation and requires endoscopic therapy
Grade 4: Active spurting, requires endoscopic therapy, possible hospital admission, and transfusion
Most bleeding episodes are clinically trivial, whereas those episodes that require hospitalization, blood transfusion, repeat endoscopic intervention, or surgery are considered true complications.
Risk Factors for Postpolypectomy Bleeding
Polyp-related factors
Risk factors include polyp size, location, and morphology. Polyp size is the main factor and is consistently linked to higher bleeding rates. Polyps that are 1 cm or larger in size have a 2.0-fold to 4.5-fold increased risk of PPB ; this risk grows by 9% to 13% per millimeter increase in polyp size greater than 1 cm. Polyps located in the right hemicolon are associated with a 2.6-fold to 4.6-fold increased risk of PPB. Polyp morphology (pedunculated with a thick stalk or laterally spreading lesions) and histology (adenoma, villous features, and presence of adenocarcinoma) are associated with an increase in PPB. In a prospective Australian Colonic Endoscopic Resection Study of patients undergoing wide field endoscopic mucosal resection (EMR) of large sessile colonic polyps, intraprocedural bleeding was associated with larger lesions, lesion histology, and Paris endoscopic classification of type 0–IIa + Is.
Patient-related factors
Patient-related factors, such as age greater than 65, hypertension, cardiac disease, and renal disease, have been associated with a higher risk for PPB. Older patients and those with cardiovascular disease are often on antithrombotic therapy when they present for colonoscopy. The use of anticoagulants before polypectomy and resumption of therapy afterward have been associated with a 3.7-fold and 5.2-fold increased risk for PPB, respectively. Risks for PPB vary based on the type of antithrombotic agent, whether an anti-platelet agent (aspirin, nonsteroidal anti-inflammatory drugs [NSAIDs], clopidogrel, ticlopidine) or an anticoagulant (warfarin, heparin), and these are reviewed later in this article ( Table 1 ).
| Antithrombotic Agent | Effect on Postpolypectomy Bleeding | Current Guidelines for Periprocedural Management a | Suggested Endoscopic Intervention |
|---|---|---|---|
| Aspirin/NSAIDs (antiplatelet) | No compelling data | Continue | Consider holding if large EMR planned |
| Clopidogrel (antiplatelet) | Mixed data: suggestion of increased delayed PPB Increases risk of PPB in combination with aspirin | Hold for 7–10 d preprocedure and restart the next day postprocedure (depending on intervention performed) | Consider prophylactic clip placement over polypectomy site(s) |
| Warfarin (anticoagulation) | Increases risk of PPB | Hold for 3–5 d preprocedure and 1 d postprocedure | |
| Heparin (anticoagulation) | Increases risk of PPB | Hold for 4–6 h (unfractionated heparin) 12–24 h (low molecular weight heparin) | |
| SSRIs (antiplatelet) | Increased risk of upper GI bleeding; no studies with colonoscopy or polypectomy Risk is further increased with simultaneous aspirin and NSAID use | N/A | Consider holding aspirin ± NSAIDs |
| Novel oral anticoagulants | Likely increased risk although no data available | No data available | Discontinue agents based on half-life and with input from cardiologist |
a ASGE guidelines: Management of antithrombotic agents for endoscopic procedures. NB: The decision to discontinue antithrombotic therapy for endoscopy should be individualized to balance the risk of the planned endoscopic intervention and with the risk of the underlying condition for which the patient is on therapy.
Aspirin and nonsteroidal anti-inflammatory drugs
Although aspirin exposure results in a prolonged bleeding time of normal colonic mucosa, studies evaluating the impact of aspirin use on PPB reveal mixed results. An increased risk for PPB with aspirin use before colonoscopy has been described in one small study (odds ratio [OR] 6.7). Another group reported increased risk of bleeding after EMR of lesions 2 cm or larger in patients with aspirin use within 7 days of the procedure (OR 6.3). However, most data show that aspirin use, alone or in combination with other NSAIDs, does not increase the risk of PPB. Variation in the aspirin dosage (81 mg vs 325 mg) may be a confounding factor.
Clopidogrel
Current guidelines recommend holding thienopyridines (clopidogrel and ticlopidine) for at least 7 days before polypectomy based on the indication for antithrombotic therapy and the risk to the patient. In a recent meta-analysis of PPB in patients on continued clopidogrel therapy, the overall pooled relative risk (RR) for PPB was 2.5 ( P <.05). The RR for delayed bleeding was 4.6 ( P <.05) with the RR for immediate PPB 1.76 (however, P >.05). Other studies have also reported higher rates of delayed PPB with uninterrupted clopidogrel therapy compared with controls, but the rates of PPB were low (less than 1%) and applied mainly to large polyps (greater than 1 cm in size). Interestingly, clopidogrel monotherapy was not identified as an independent risk factor for PPB, but only increased the risk of PPB when used in combination with aspirin. The risk of delayed PPB can be mitigated during colonoscopy by the application of through-the-scope clips (TTSC; endoclips) to the polypectomy site in patients who are on clopidogrel therapy during colonoscopy or are expected to resume it after the procedure.
Warfarin
It is recommended that patients anticoagulated with warfarin should hold therapy for 3 to 5 days before colonoscopy. Although there is no clearly established target international normalized ratio (INR) that is safe for polypectomy, an INR of 1.6 or less is generally considered acceptable. However, preprocedure INR is not routinely obtained in clinical practice. Some studies showed that patients on warfarin therapy had a higher risk of PPB (OR 11.6–13.3), whereas those on aspirin, NSAIDs, thienopyridines, and low-molecular-weight heparin did not. Other predictors of PPB were the number of polyps removed (OR 1.2) and male gender (OR 9.2).
Bridging therapy with heparin
When evaluating the risk of warfarin therapy on PPB, it is important to consider the risk of thromboembolic events after temporary discontinuation of the drug. Patients at high risk for thromboembolic events are bridged with heparin while anticoagulation is suspended. In a CORI database study, the use of periprocedural heparin was not a risk factor for PPB. In another study, patients on warfarin or antiplatelet agents were either bridged with heparin or simply discontinued antithrombotic therapy. In patients on heparin bridge therapy, the risk of PPB was not only higher but also delayed in onset and recurrent, resulting in prolonged hospitalization. Warfarin use, bridging with heparin, and pedunculated polyps were independent predictors of PPB. Endoscopists should consider empiric placement of endoclips at polypectomy sites in patients on warfarin therapy. Patients with an average INR of 2.3 and polyps ranging from 3 mm to 10 mm in size were successfully clipped and experienced no PPB when warfarin was held for only 36 hours preprocedure.
Thrombocytopenia and selective serotonin reuptake inhibitors
Endoscopists often hesitate to perform polypectomy on patients with cancer with thrombocytopenia or cirrhosis of the liver because of a perceived increased risk of PPB. In a study of patients with cancer with platelet counts of 75 × 10 3 or less, the risk of bleeding was 1.5% after biopsy and 4% after polypectomy for lesions 1 cm in size or smaller. In addition, hemostasis was successfully achieved in 95% of actively bleeding patients, resulting in a decreased need for transfusion of blood products.
Patients with early-stage liver cirrhosis (Child-Pugh score A or B) with an INR of 1.5 or less and platelet counts ranging between 30 × 10 3 and 242 × 10 3 had immediate PPB in 3% of patients; immediate hemostasis was achieved with clips, and there was no delayed bleeding. Biopsy and polypectomy may be safe in patients with thrombocytopenia or early-stage cirrhosis with appropriate precautions. Patients with more advanced cirrhosis (Child-Pugh score B or C) were at increased risk for PPB (hazard ratio [HR] 3.5) along with polyp size (HR 3.6) and pedunculated type polyps (HR 2.4).
The use of selective serotonin reuptake inhibitors (SSRIs) has been associated with an increased risk of both organ hemorrhage and adverse outcomes of surgery. SSRIs are thought to inhibit platelet adhesion by up to 50%, an effect that varies by agent and dosage. Multiple studies have demonstrated an increased risk of upper gastrointestinal (GI) bleeding in patients using SSRIs (OR 1.6). This effect is compounded by the concomitant use of NSAIDs (OR 4.2–8.0) and even more so when aspirin is added to the mix (OR 28.0). There are little data evaluating the effect of SSRIs on lower GI bleeding to make recommendations at this time.
Technique/Device-Related Factors
There is wide variation in endoscopic resection techniques among gastroenterologists, specifically with the type of cut (cold vs hot snare resection) and type of current (pure cut vs pure coagulation vs blended current).
Polyps less than 1 cm in size are often removed with a biopsy forceps or a cold snare. Although “cold” snare resection of polyps increases risk of intraprocedural bleeding, it is easily controlled; it is not associated with delayed PPB as observed in a prospective multicenter trial of more than 1000 polypectomies. There is no difference in PPB between cold-snare and hot-snare resection of polyps less than 1 cm. However, rates of PPB were higher in hot-snare polypectomy in patients on warfarin.
Cautery is generally used to resect polyps greater than 1 cm in size. There is variability among gastroenterologists in the electrosurgical settings used for snare resection in the United States: 46% use pure coagulation, 46% use blended current, and 3% use pure cut for polypectomy. Pure cutting current increases the risk of bleeding, while pure coagulation current increases the risk of perforation. Hence, a blended cutting and coagulation current is used to minimize both of these complications. Although there was no difference in the rate of PPB, immediate bleeding was more common after blended current use, whereas delayed bleeding was frequent in anticoagulated patients. This area needs further investigation to help standardize endoscopic practice.
Prevention of Postpolypectomy Bleeding
Several studies report outcomes of the different techniques available to prevent PPB. These techniques include injectable solutions, placement of endoclips, deployment of a detachable loop, and application of thermal energy (with coagulation forceps or argon plasma coagulation) as well as a combination of these techniques ( Box 2 ).
Early bleeding
Epinephrine injection superior to control (OR 0.37)
Combination therapy superior to monotherapy (OR 0.12)
Detachable loop superior to epinephrine or control (OR 0.25)
Delayed bleeding
Epinephrine versus control: no difference in bleeding rates
Combination therapy versus monotherapy: no difference in bleeding rates
Neither detachable loops nor endoclips versus other techniques: no difference in bleeding rates
Solutions—epinephrine injection
Submucosal injection of fluid underneath a polyp elevates the lesion, separates it from the deeper layers, prevents transmural thermal injury, and thus improves the safety of polypectomy. Dilute epinephrine offers the additional advantage of shrinking polyp size and reducing the risk of bleeding. This principle has been used to remove large pedunculated polyps with the injection of dilute epinephrine into the head and into the stalk of the polyp for volume reduction (epinephrine volume reduction). Compared with saline injection or no preinjection, epinephrine injection results in lower PPB. Saline tends to dissipate quickly from the submucosal cushion in highly vascularized areas like the rectum. Although this could be minimized by hypertonic saline (3.0%–4.7%) in combination with epinephrine or 50% dextrose, these solutions can result in local inflammation and delayed tissue damage. A 0.5% hyaluronic acid solution provides the most durable cushion but is also associated with an inflammatory reaction. A combination of saline and methylene blue or indigo carmine dye stains the submucosal space and provides visual confirmation of the depth of resection. Use of other solutions have been reported (fibrinogen, albumin, glycerol), but these are expensive and there are little data to support their use.
Endoclips
Endoclips are frequently used to prevent PPB. Although clips have not been shown to be useful in the prevention of PPB after resection of smaller lesions, a recent study demonstrated that clips prevent delayed bleeding after EMR of lesions greater than 2 cm. When endoclips were placed before polypectomy in patients on uninterrupted antithrombotic therapy, there was no PPB in either the endoclip or the control groups. It is cost-effective in patients planning to resume antithrombotic therapy after the procedure, as shown in a decision analysis. Clip application may be ineffective if the clip is unable to span across a thick stalk and occlude all the feeding blood vessels and may cause counterburn during snare resection if applied to a short stalk.
Detachable loops
Detachable loops applied to the stalk of a pedunculated polyp result in vasoconstriction of feeding blood vessels and reduce the risk of PPB. The efficacy of endoloops in the prevention of PPB is similar to other mechanical hemostatic devices such as clips. Endoloop ligation is particularly useful in the prevention of PPB from large polyps (>2 cm).
Cautery
Cautery can be applied to treat and prevent PPB. However, neither argon plasma coagulation nor coagulation forceps treatment of nonbleeding vessels has been effective in the prevention of PPB.
Combination therapy
Combination hemostatic therapy with epinephrine injection and application of mechanical devices (detachable loop or endoclip or both) reduces rates of PPB as well as severity of PPB in large polyps (>2 cm) compared with epinephrine injection alone.
A meta-analysis and systematic review summarizes the data from available quality studies. Both monotherapy (epinephrine injection or mechanical hemostasis) and a combination of epinephrine injection and mechanical hemostasis reduce early PPB; combination therapy is superior to monotherapy in the prevention of early PPB. Epinephrine injection alone is ineffective in the prevention of delayed PPB, whereas monotherapy with mechanical devices and a combination of epinephrine and mechanical devices are effective in the prevention of delayed PPB.
Management of Postpolypectomy Bleeding
Three options are currently available to manage PPB depending on the timing and severity of bleeding: injectable solutions, thermal devices, and mechanical devices. A decision of whether to treat bleeding during or immediately after resection can be made and implemented promptly ( Box 3 ).
Thermal (contact)
- •
Heater probe
- •
Gold probe
- •
Snare tip cautery
- •
Hemostatic forceps
- •
Thermal (noncontact)
- •
Argon plasma coagulation
- •
Mechanical
- •
TTSC or endoclips
- •
Detachable loop
- •
OTSC
- •
Grasp bleeding stalk with snare device and constrict
- •
Band ligation
- •
Endoscopic suturing device
- •
Intraprocedural bleeding is associated with clinically significant delayed PPB after wide field EMR, and it should be treated aggressively and immediately. One could consider injection of epinephrine or application of clips, loops, or cautery to control bleeding immediately depending on the source of bleeding.
Delayed PPB requires careful assessment of the patient and the severity of bleeding. The best course of action should be determined based on clinical features, such as hemodynamic instability, frequency of bloody bowel movements, comorbid conditions, patient age, resumption of antithrombotic agents, and so on. If resuscitation measures fail, patients should undergo urgent intervention: colonoscopy with hemostasis, angioembolization, or surgery. If resuscitation measures are successful, patients should be admitted to hospital for observation and monitoring of hemoglobin, platelets, and coagulation profile. More than half of all bleeding subsides spontaneously and does not need any further intervention. Risk factors for ongoing bleeding or recurrence of bleeding that requires intervention include hematochezia hourly or every few minutes, hemodynamic instability, low hemoglobin on admission (<12 g/dL), transfusion requirement, and American Society of Anesthesiologists class of II or higher. If bleeding resolves spontaneously, the patients should be observed for 24 hours before discharge. If it continues or recurs, urgent intervention with colonoscopy is required.
Injectable solutions for hemostasis
In the setting of active bleeding, injection of a 1:10,000 epinephrine-saline solution into the site will help achieve temporary hemostasis by a tamponade effect as well as vasoconstriction of the bleeding vessel.
Thermal hemostatic devices
Thermal devices generate heat and cause edema, coagulation, and vessel constriction through direct contact with the tissue (heater probe, gold probe, snare tip, hemostatic forceps) or indirectly (argon plasma coagulation). Bipolar devices, such as the heater probe and gold probe, complete the circuit within the device and no grounding pad is needed. Although tissue contact is necessary, pressure should not be applied in the colon, unlike when these devices are used in the stomach, because the colon wall is thinner and the settings should also be reduced accordingly (15 J heater probe, 10–15 W gold probe).
When bleeding occurs after polypectomy, the snare is often readily available and the tip of snare can be applied to the bleeding site and soft coagulation applied. The snare is extended out of the sheath by 1 to 2 mm and the wire tip is applied to the bleeding point. Soft coagulation is applied until hemostasis occurs. The coagulation setting allows for desiccation of the tissue, which creates resistance to further current flow, thus limiting injury. This method alone (80 W, effect 4) was successful in 91% of cases when immediate PPB was seen. Studies have also demonstrated the use of hemostatic forceps or graspers in pinching and retracting bleeding points to stop postpolypectomy and diverticular bleeding.
Noncontact thermal energy can be applied with argon plasma coagulation where electron flow through ionized argon gas results in tissue desiccation. Again, desiccated tissue limits further energy flow, and the argon stream will shift to adjacent conductive tissue, allowing for hemostasis of the bleeding vessel.
Mechanical hemostatic devices
The most popular method for endoscopic hemostasis is the application of endoclips or TTSC. In a survey of United States Department of Veteran Affairs gastroenterologists, placement of an endoclip was the preferred method to achieve hemostasis in cases of PPB in 76% of respondents (as well as for bleeding prophylaxis in patients on anticoagulation). For persistent bleeding after cold forceps biopsy, 55% of gastroenterologists would not apply any therapy. Multiple studies have demonstrated that application of TTSC to the bleeding vessel achieves successful hemostasis. A novel technique of endoscopic clip tamponade has been described for bleeding in the midst of an EMR. The endoclip is closed over the lesion for 3 minutes to control bleeding; then the clip is opened so that resection may continue. TTSC placement has also been used in combination with a detachable loop, which is placed to lasso the deployed endoclips together and provide an additional level of mechanical force.
Other devices for mechanical hemostasis include the detachable loop, which is placed more easily over protuberant lesions like a bleeding polypectomy stalk. Endoscopic band ligation can be used in these situations. Although over-the-scope clips (OTSC) play a greater role in closing perforations, they been effective in PPB in small studies. The precise placement of an OTSC may be challenging in the face of an active arterial bleed and would also require the scope to be withdrawn from the patient after localizing the site to retrofit the clip on to the scope. There have been no comparative studies of TTSC with OTSC for hemostasis in humans. The endoscopist should use the device that they are most familiar with and that is readily available in the setting of an acute arterial PPB.
Colon perforation
Colon perforation, a dreaded complication of polypectomy, is rare with routine polyp removal. The incidence of colonoscopic perforation varies from 0.7% to 0.9%. However, the incidence of perforation has certainly increased with the introduction of endoscopic submucosal dissection (ESD) in the colon. The rates of perforation with ESD are 4% to 10%.
The colon wall is approximately 3 mm thick. The submucosa is the strongest layer in the GI tract. Full-thickness resection of the submucosa leaving the muscularis propria intact results in postpolypectomy syndrome and delayed perforation ; this could be avoided by prophylactic clip closure of deep resections where the muscularis propria is exposed.
Once a perforation occurs, air escapes into the peritoneum. A massive air leak could result in tension pneumoperitoneum and cardiovascular arrest ; this could be prevented by the routine use of carbon dioxide instead of room air for colon insufflation, because carbon dioxide gets reabsorbed into the body faster than room air. In addition, periodic decompression of the colon by removing the biopsy cap and allowing the gas in the colon to vent out reduces this risk. Tension pneumoperitoneum presents with pulseless electrical activity. Once recognized, this requires immediate treatment. It involves insertion of a wide-bore needle or angiocath into the peritoneal cavity after confirming the presence of pneumoperitoneum by insertion of a smaller-gauge needle attached to a syringe filled with saline and observing air bubbles come up the syringe. One could leave the endoscope in the colon as abdominal decompression is performed, so that the perforation could be closed once the patient is stabilized.
Within minutes of perforation, fluid leaks out of the colon and peritonitis sets in. If stool escapes, fecal peritonitis sets in. Hence, it is important to have a clean colon and aim for an excellent colon preparation. In addition, making an effort to suction and dry up the colon segment where a resection is performed as well segments proximal and distal to the site of resection avoids the risk of flooding the site of resection if colon perforation were to occur. Placing the lesion in a nondependent position minimizes the risk of flooding of fluid and avoids fluid escape in case of perforation.
Once a perforation occurs, one should be ready to close it immediately if needed. Small perforations without peritoneal leakage of gas or fluid can be closed after extending and completing the resection; hurried closure and entrapment of polyp in the clip closed defect may result in a refractory recurrence that may be difficult to eradicate. Closure can be undertaken with TTSC or OTSC.
Microperforation and Macroperforation
Colonic perforations can be categorized based on the size of perforation and the timing of detection. Microperforations occur because of colonic wall thinning after the application of cautery and dissection of the submucosal layer that is not detected during endoscopy. These microperforations are detected when free air is seen on imaging studies done routinely after ESD in some centers. Macroperforations occur because of deep and full-thickness resection of the colon wall and are obvious to the endoscopist in real time. These macroperforations result from inadequate injection of submucosal solution that fails to lift the lesion adequately, resection of the lesion into deep submucosa, or submucosal fibrosis from prior manipulation of the lesion (ie, attempted polypectomy).
Although immediate perforation detected during colonoscopy may be addressed endoscopically, delayed perforation often requires surgery. Patients presenting within 24 hours after the procedure are likely to have minimal peritoneal contamination and can successfully undergo surgical repair compared to patients with delayed presentation where fecal peritonitis has occurred and require a diversion procedure. Risk factors for postoperative morbidity include blunt injury, poor bowel preparation, corticosteroid use, and age less than 67.
Risk Factors for Perforation
Polyp-related factors
Polyp-related risk factors include location, size, and morphology of the lesion. Cecal location is an important risk factor for perforation. A nonpedunculated polyp in the cecum is associated with a 12-fold increased risk of perforation over a similar sized polyp in the distal colon. Nonpolypoid or laterally spreading lesions are at risk for perforation (OR 4.1). Lesions that may involve deeper tissue layers (Vienna classification 4: noninvasive high-grade dysplasia and Vienna classification 5: invasive neoplasia) were an independent risk factor for perforation. Submucosal fibrosis from prior polypectomy attempts will impair successful lifting of the lesion with submucosal injection and increase the risk of perforation (OR 4.5). Submucosal injection of hyaluronic acid solution is protective against perforation because of its ability to prolong the submucosal cushion effect for several minutes (OR 0.3). It is critical to evaluate the morphology of the lesion in terms of their risk for perforation when contemplating resection so that adequate precautions can be taken ( Box 4 ).
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
