Construction and validation of a risk prediction model for perianal infection in patients with haematological malignancies during chemotherapy: a prospective study in a tertiary hospital in China.

STRENGTHS AND LIMITATIONS OF THIS STUDY

Data were prospectively collected from tertiary hospitals with a high number of haematological malignancies.

The risk factors included in the model were available in the hospitals’ information systems at all levels.

This model facilitates dynamic assessments and advances drug and non-drug preventive interventions.

The influence of chemotherapy was not considered when developing the prediction model.

Additionally, the suitability of this model for other institutions requires further external validation.

Introduction

Infection is an important complication in patients with haematological malignancies during chemotherapy due to a compromised immune system. 1 The perianal area is a common site of infection with an incidence of 9.2%–27%, which is higher than that of other cancers. 2 This is possible due to more severe bone marrow suppression and immunocompromised status in patients with haematological malignancies. Perianal infection is often accompanied by severe pain, swelling and constipation, which have a serious impact on the patients’ quality of life. 3 In severe cases, sepsis may occur and endanger the patient’s life, with the mortality of patients with perianal infection being as high as 16%. 4 Even after treatment, the recurrence rate is 31%. 5

Therefore, early prediction and aggressive prevention of perianal infection during chemotherapy are critical in reducing the incidence of perianal infection, thereby decreasing the morbidity and mortality due to chemotherapy, elevating the quality of life and improving prognosis. 6

However, most existing studies have focused on the diagnosis and treatment after perianal infection has occurred 7 8 and no risk prediction model has been established. Therefore, to help medical staff promptly screen patients at a high risk of perianal infection during chemotherapy, this study comprehensively identified risk factors for perianal infection. We developed and validated a clinical prediction model for the probability of perianal infection. To the best of our knowledge, this is the first model to predict perianal infection during chemotherapy in patients with haematological malignancies.

Methods

Patient and public involvement statement

The patients and/or the public were not involved in the design, conduct, reporting or dissemination of this study.

Patients

This was a single-centre, prospective, observational study. Convenience sampling was used to recruit patients with haematological malignancies during chemotherapy at the department of haematology, West China Hospital of Sichuan University between January 2021 and June 2022.

The included patients met all the following criteria: (A) patients with a confirmed diagnosis of haematological malignancies; (B) no less than 18 years of age; (C) no perianal infection before chemotherapy; (D) no other accompanying perianal diseases, such as anal fistula or anal polyps and (E) having provided informed consent. The exclusion criteria were as follows: (A) haematopoietic stem cell transplantation, (B) discharge or death within 1 week after chemotherapy and (C) missing important clinical data (eg, incomplete medical history or missing daily stool registration). The enrolled patients were divided into training (n=595) and validation (n=255) groups in a 7:3 ratio.

Data collection

Through a literature review, small-scale expert consultation, and our pilot study of perianal infection in patients with haematological malignancies, we collected the following candidate variables that can be assessed early on hospital admission: (A) demographic data included gender, age, marital status and education level; (B) clinical characteristics composed of body mass index, diagnosis, disease stage; (C) whether presenting with diarrhoea or constipation and the duration of symptoms during chemotherapy; (D) history of diabetes, haemorrhoids, anal fissures or perianal infection and (E) laboratory examinations including serum albumin, haemoglobin, white blood cells and neutrophils within 7 days after chemotherapy. If more than one laboratory test result was available, the highest values of C reactive protein, procalcitonin (PCT) and interleukin-6 were used, and the average was calculated for other items.

Quality control

Before the initiation of the study, the members of the research group were trained to familiarise themselves with the research content to ensure consistency in data collection.

The researchers collected information on the demographic characteristics, clinical characteristics and medical history of the patients through on-site interviews within 24 hours of admission. Laboratory examinations were performed within 7 days of chemotherapy. Within 1 week of discharge, the collected data were checked against the electronic medical record system by two haematology nurses with the same qualifications. The consistency of results between the two nurses was verified. Discrepancies were resolved through discussion and verification with a data collector. If the two data verifiers disagreed with the verification results, the haematologists were invited to participate in the adjudication of accurate data. The data collected in this study were highly consistent with 100% agreement on whether perianal infection occurred in each case. The final verified data were entered into an Excel spreadsheet by the two nurses.

Outcome measure

The outcome measure was the occurrence of perianal infection at the time of discharge. The diagnosis of perianal infection was made jointly by a haematologist and a physician specialising in anorectal disease. The evaluation criteria for perianal infection were as follows: degree I, localised inflammation such as redness, swelling, heat and pain; degree II, localised inflammation with abscess formation and degree III, skin ulceration, necrosis, bleeding, sinus or fistula formation. 9 In this study, any degree of infection after chemotherapy was considered a perianal infection.

Statistical analyses

The sample size was calculated as 10 times the number of independent variables in the prediction model. 10 Assuming that each variable required at least 10 cases for verification and 10% of the incomplete data; the expected sample size should be at least 267 cases. Multiple imputations were used to handle missing data. 11 Patients with more than 30% missing data were excluded.

Variables with a variance inflation factor (VIF) >10 were not included in the subsequent analysis for multicollinearity. 12 Categorical variables are expressed as frequency (%). Continuous variables are expressed as mean±SD if they conformed to a normal distribution or median and IQR if they deviated from a normal distribution. An independent sample t-test or Wilcoxon Mann-Whitney U rank-sum test was used to analyse differences in continuous data. Categorical data were compared using the χ 2 test or Fisher’s exact test. Variables with significant differences (p<0.05) were incorporated into the multivariate logistic analysis to screen for independent predictors. 13 A nomogram prediction model was constructed to transform the regression equations into visual graphics, making the predictive model readable and convenient for clinical applications.

The ROC curve was plotted based on the predictions of the nomogram model. The capability of discrimination was evaluated using the area under the curve (AUC) of the ROC curve, where 0.5<AUC<0.7 indicates that the model’s distinguishing ability is low; 0.7<AUC<0.9 indicates a moderate ability and AUC>0.9 reflects a high discrimination ability. 14 The Hosmer-Lemeshow goodness-of-fit test was used to evaluate the calibration of the prediction model. A p>0.05 was considered to indicate no significant difference between the predicted value of the model and the actual observation. 15 We assessed the net benefit using decision curve analysis and clinical usability using clinical impact curve analysis.

All analyses were performed using the SPSS (V.25.0; IBM) and R Statistical Software (V.4.0.2; R Core Team 2020).

Results

Patient inclusion

A total of 897 patients with haematological malignancies undergoing chemotherapy were initially screened, and 47 were excluded because they did not meet the following inclusion and exclusion criteria: 28 patients were discharged from the hospital within 2 weeks after chemotherapy, 3 patients died within 1 week after chemotherapy, 4 patients had perianal infection before chemotherapy, 3 patients had perianal complications and 9 patients had missing important clinical data. A total of 850 patients who met the inclusion criteria were included. The incidence of perianal infection in this cohort was 10.5% (89/850 patients). These cases were divided into training (595 cases) and validation (255 cases) groups ( table 1 , figure 1 ). Differences in demographic data, clinical characteristics, laboratory indicators and other data between the two groups were not statistically significant (p>0.05), indicating that the two groups were comparable.

Development of the predictive nomogram

No apparent multicollinearity was observed among the 24 variables in the training group (VIF<10). In the univariate analysis, there were significant differences in the history of perianal infection, haemorrhoid history, anal fusion history, duration of diarrhoea, constipation, PCT and white cell count (WCC) between the perianal infection and non-infection groups (p<0.05, online supplemental table 1 ). Logistic multivariate regression analysis suggested that a history of perianal infection, haemorrhoids, constipation and diarrhoea duration were independent risk factors for perianal infection (p<0.05, figure 2 ).

bmjopen-2023-074196supp001.pdf

The four independent risk factors were entered into the R software, and the ‘lrm’ function in the ‘rms’ package was used to build the nomogram prediction model ( figure 3 ). The value of each factor was determined based on the intersection of the vertical line drawn from the variable to the point axis, and the total risk score was calculated by adding all the variable points. The probability of perianal infection was determined directly from the total point axis.

Validation of the prediction model

The area under the ROC curve was 0.784 (95% CI 0.727 to 0.841) in the training group and 0.789 (95% CI 0.818 to 0.885) in the validation group ( figure 4A , online supplemental figure 1A ). In the training group, at a cut-off value of 10.6%, the sensitivity and specificity were 81.8% and 70.5%, respectively. In the validation group, at a cut-off value of 5.7%, the sensitivity and specificity were 91.3% and 65.2%, respectively. The model also had good calibration in both the training and validation groups, with a non-significant Hosmer-Lemeshow test (p=0.999 and 0.482, respectively). The calibration curve showed good consistency between the predictions and actual observations in both the training and verification groups ( figure 4B , online supplemental figure 1B ).

bmjopen-2023-074196supp002.pdf

The decision curves for the training and verification groups are shown in figure 4C and online supplemental figure 1C . The abscissa represents the threshold probability, and the probability of perianal infection in patients with haematological malignancies during chemotherapy as predicted by the model is represented as Pi. When Pi reached a certain threshold (recorded as Pt), it was defined as positive, indicating that the model benefited the patient (gain). However, there will also be losses in patients with unpredicted perianal infection (loss), and the net benefit (NB) is calculated as gain minus loss. In the training group, the nomogram had the highest NB within the threshold probability range of 10%–60%. Similar results were observed in the validation group. The clinical impact curves for the training and verification groups are shown in figure 4D and online supplemental figure 1D .

The clinical impact curve showed that the model could be further used to predict the risk of perianal infection in 1000 patients with haematological malignancies undergoing chemotherapy. The black curve indicates the number of positive cases (high risk of perianal infection) screened using the constructed model and the blue curve indicates the number of true positives (actual occurrence of perianal infection). Under each threshold probability, when the threshold probability of high perianal infection risk is ≥0.4, the prediction model can obtain a high degree of coincidence between the theoretical and actual events and has high clinical validity.

Discussion

To the best of our knowledge, there are few studies on the risk factors of perianal infection in patients with haematological malignancies, and the results remain controversial. This study is the first to develop and evaluate a predictive nomogram for the risk of perianal infection during chemotherapy in patients with haematological malignancies. This nomogram was developed and validated in patients from a high-volume tertiary medical institution. The measurements of the four parameters in the predictive nomogram can be obtained simply and non-invasively, facilitating widespread ease of adoption in clinical practice. In the discrimination test, the area under the ROC curve of the prediction model in the training group was 0.784, and the area under the ROC curve in the validation group was 0.789, with a difference of only 0.005, indicating that the model had a strong predictive ability and high accuracy in identifying perianal infection during chemotherapy in patients with haematological malignancies. In the calibration test, the p values were 0.999 and 0.482, respectively, which were much higher than 0.05, indicating that the probability of perianal infection predicted by the model was close to the actual incidence, and that the calibration of the prediction model was good. In the evaluation of clinical validity, when the threshold probability of high perianal infection risk is ≥0.4, the prediction model can obtain a high degree of consistency between the theoretical and actual events, indicating high clinical validity. Therefore, the results of this study showed that the model could be used to screen patients at a high risk of perianal infection and make individualised clinical decisions based on the prediction results of the model. This model can help achieve early prediction, prevention and intervention to reduce the incidence and mortality of perianal infection, reduce the length of hospital stay and hospitalisation expenses, and improve the patients’ long-term quality of life.

Notably, four independent predictors were identified. A history of haemorrhoids was considered the most important predictor. Patients with a history of haemorrhoids had more than five times the probability of perianal infection during chemotherapy (OR=5.369). Haemorrhoids are soft clusters of veins formed by the flexion and expansion of the subcutaneous venous plexus of the anal canal. 16 In 2016, the incidence of anorectal diseases among urban residents in China was 51.14%, among which the incidence of haemorrhoids was the highest, accounting for 50.28% of all anorectal diseases. 17 In this study, 38.35% (326/850) of patients had a history of haemorrhoids. Previous studies have shown that such patients have a 7.733 times greater risk of perianal infection than those without a history of haemorrhoids. 18 This may be because during chemotherapy, there is more treatment fluid; owing to fatigue after chemotherapy, patients have more bed rest, bowel habits change, patients face difficulty defecating, abdominal pressure increases, the original haemorrhoids rupture and bleed, providing opportunities for bacterial invasion, which increases the risk of infection. 19 In patients with long-term diarrhoea, the anus and rectum are frequently stimulated, causing phlebitis and periphlebitis of the venous plexus, leading to vascular wall embrittlement, secondary vascular dilatation and congestion, 20 which in turn increases the risk of infection. Medical management (eg, stool softeners, topical over-the-counter preparations, topical nitroglycerine), dietary modifications (eg, increased fibre and water intake) and behavioural therapies (sitz baths) are the mainstays of initial therapy. 21 In severe cases, surgery can be performed after the recovery of myelosuppression during chemotherapy.

Patients with a history of perianal infection have a higher probability of reinfection than those without perianal infection, regardless of treatment success. 22 Mo et al 23 found that a history of perianal infection was an independent risk factor for perianal infection, which was consistent with the results of this study. In this study, 27.14% (34/124) of the patients with perianal infection had a history of perianal infection. In patients with perianal infection, the local skin of the lesion remains abnormal even after curing. Additionally, there may be local drug-resistant bacteria, and perianal infection is very likely to relapse during the neutropenia stage during chemotherapy. Therefore, patients with haematological malignancies and a history of perianal infection should pay more attention to perianal hygiene during chemotherapy. Perianal cleaning should be performed after urination. In the morning and evening, sitz baths can also be performed with povidone-iodine diluent, metronidazole plus saline and traditional Chinese medicine preparations 24–26 to prevent the recurrence of the infection during the myelosuppression period.

Diarrhoea in patients with haematological malignancies can be divided into antibiotic-associated diarrhoea, chemotherapy-induced diarrhoea and infection-associated diarrhoea. In patients with haematological malignancies undergoing chemotherapy, the toxicity and side effects of chemotherapeutic drugs can seriously damage the gastrointestinal mucosa, leading to gastrointestinal mucositis. Villus damage reduces the surface area available for absorption in the intestine and reduces the water absorption rate. 27 Simultaneously, in the myelosuppression stage, neutrophils decrease and intestinal infection becomes imminent. 28 The use of antibiotics and the disorder of intestinal flora may make patients develop diarrhoea. 29 Previous studies have found that the incidence of diarrhoea in patients with haematological malignancies during chemotherapy is as high as 74.2%. The incidence of diarrhoea in this study was 19.06% (162/850), which was lower than that reported in previous studies; however, 28.40% (46/162) of them had severe diarrhoea for more than 3 days. Normal stool is neutral, weakly acidic or weakly alkaline, whereas diarrhoea stool is strongly acidic or strongly alkaline, containing pancreatic lipase, bile salts and ammonia base to stimulate perianal local mucosa, causing inflammation and injury. 30 With prolonged duration of diarrhoea, the perianal skin is continuously polluted and stimulated by faeces, becoming continuously moist; repeated perianal wiping is more likely to cause perianal skin damage and erosion, and increase the incidence of infection. Therefore, patients undergoing chemotherapy should provide increased importance to environmental and dietary hygiene and pay close attention to their stool status. If the frequency of stools increases or the character changes, active drug control should be performed. A soft baby paper towel can be used after urination; the perianal area can be washed with a cotton cloth dipped in warm water, and the perianal skin can be dried with a blow dryer to keep it dry, if necessary. If there is a local inflammatory reaction, products such as skin protective films or powders can be used to effectively isolate faeces, protect the skin, promote healing and avoid the aggravation of infection.

Gastrointestinal reactions caused by chemotherapy drugs, the application of antiemetic and antacid drugs, changes in diet structure and other factors may cause constipation in patients with haematological malignancies during chemotherapy. 31

Faecal retention in the intestine and excessive absorption of water may result in dry stool, dry and hard stool, and perianal mucosa contact, which can easily cause mechanical damage to the perianal mucosa. Stool accumulation for an extensive time may cause the blockage of perianal venous return, which affects the local blood circulation of the perianus, perianal swelling and local anal fissure; it is an important cause of perianal infection. In this study, 32.35% (275/850) of the patients had constipation and 48.36% (133/275) had constipation for more than 3 days, indicating that the incidence and duration of constipation were high in this study. This indicates that if constipation occurs and is left untreated, it can cause perianal infections. The duration of constipation does not affect the occurrence of perianal infections if the clinician provides treatment to soften the stool and help defecation. During chemotherapy, patients with haematological malignancies should have a healthy diet, eat more fresh fruits and vegetables, and use enemas, if necessary, to ensure unblocked stool, avoid local mucosal damage and prevent perianal infections.

The extent and duration of leucopenia and neutropenia in relation to perianal infection has always been a hotspot of concern. Previous studies have shown controversy regarding whether leucopenia, lack of granulocytes and duration of granulocyte deficiency are risk factors for perianal infections. Shtyrkova et al 32 and Zheng et al 33 believe that granulocyte deficiency is an independent risk factor for perianal infection, as the lack of granulocytes leads to a natural defence against infectious microorganisms. The number of granulocytes determines the form of inflammation, the process of infection and the chance of developing sepsis. In contrast, Orhan et al argued that there is no relationship between postchemotherapy WCC, neutrophil count, duration of neutrophil deficiency and the occurrence of perianal infection. 2 Although patients commonly experience bone marrow suppression after chemotherapy, they are more focused on personal protection during periods of granulocyte deficiency, and doctors often use antibiotics empirically to prevent infection, thereby reducing the incidence of infection. Our results show that the postchemotherapy WCC, neutrophil count and duration of neutrophil deficiency are not independent risk factors for perianal infection, further proving that the main factors influencing the occurrence of perianal infection during the chemotherapy period in patients with haematological malignancies are local factors in the perianal region.

Nevertheless, this study has several limitations: first, the diagnosis of perianal infection was not the result of objective examination, such as ultrasound or CT examination. In this study, the diagnosis of perianal infection was mainly made by haematologists and physicians specialising in haemorrhoids and fistulas based on clinical experience. Second, previous studies have reported that chemotherapy regimen and intensity have a great impact on the duration of bone marrow suppression 34 ; therefore, it could be one of the risk factors for perianal infection in patients. However, various kinds of diseases and complex chemotherapy regimens in patients with haematological malignancies were included in this study, making it difficult to perform stratified analysis. Finally, because of the limited experimental conditions and other resources, this study is a single-centre study, and the model has not been externally verified by other centres. Therefore, in future studies, the diagnosis of perianal infections should be determined using objective examination results. At the same time, the sample size was expanded, more influencing factors were included, and confounding factors such as chemotherapy regimen and chemotherapy intensity were controlled for stratified analysis. The predictive value of the nomogram model needs to be verified via a multicentre study.

Conclusions

This study found that a history of perianal infection, haemorrhoids, constipation or diarrhoea for >3 days could increase the incidence of perianal infection in patients with haematological malignancies during chemotherapy. In other words, patients with haematological malignancies, underlying perianal diseases and abnormal defecation have a higher risk of perianal infections during chemotherapy. The risk prediction model constructed in this study can be helpful in guiding clinical medical staff in screening for early intervention in high-risk groups of perianal infections in patients with haematological malignancies during chemotherapy with satisfactory accuracy.

Supplementary Material

Reviewer comments

Author's manuscript

Acknowledgments

We are grateful to the Clinical Research Management Department of West China Hospital of Sichuan University for providing guidance on the topic design and statistical analysis of this study. Thanks to all patients who provided clinical data for this article.

Footnotes

Contributors: TN obtained funding for the project, designed the study responsible for the overall content as the guarantor. YW and YLuo performed the experiments and drafted the manuscript. YW, YLuo, YLeng and MY performed this study. YLuo and TL revised the article. YW analysed the data, and prepared figures and tables. The authors read and approved the final manuscript.

Funding: This work was supported by 1.3.5 Project for Disciplines of Excellence, West China Hospital, Sichuan University (No. ZYJC21007) and the West China Nursing Discipline Development Special Fund Project, Sichuan University (HXHL21024).

Competing interests: None declared.

Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review: Not commissioned; externally peer reviewed.

Supplemental material: This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants and the trial was approved by the Ethics Committee on Biomedical Research at West China Hospital of Sichuan University (approval code: NO. 2020-1170). All procedures performed in this study involving human participants were following the Declaration of Helsinki (as revised in 2013). Informed consent was obtained from all patients. Participants gave informed consent to participate in the study before taking part.

References

1. Rieger CT, Liss B, Mellinghoff S, et al. Anti-infective vaccination strategies in patients with hematologic malignancies or solid tumors-guideline of the infectious diseases working party (AGIHO) of the German society for hematology and medical oncology (DGHO). Ann Oncol 2018;29:1354–65. 10.1093/annonc/mdy117

2. Orhan B, Özkalemkaş F, Özkocaman V, et al. The role of white blood cell count in perianal pathologies: a retrospective analysis of hematologic malignancies. Mediterr J Hematol Infect Dis 2022;14:e2022051. 10.4084/MJHID.2022.051

3. Solmaz S, Korur A, Gereklioğlu Ç, et al. Anorectal complications during neutropenic period in patients with hematologic diseases. Mediterr J Hematol Infect Dis 2016;8:e2016019. 10.4084/MJHID.2016.019

4. Baker B, Al-Salman M, Daoud F. Management of acute perianal sepsis in neutropenic patients with hematological malignancy. Tech Coloproctol 2014;18:327–33. 10.1007/s10151-013-1082-z

5. Chen C-Y, Cheng A, Huang S-Y, et al. Clinical and microbiological characteristics of perianal infections in adult patients with acute leukemia. PLoS ONE 2013;8:e60624. 10.1371/journal.pone.0060624

6. Jiang Q, Zhang D, Majaw J, et al. Minimization of the perianal infection rate of hematological malignancies with agranulocytosis by quality control circle activity and patient-hospital-student win-win concept. J Int Med Res 2018;46:2338–45. 10.1177/0300060517726863

7. Haliloglu N, Gulpinar B, Ozkavukcu E, et al. Typical MR imaging findings of perianal infections in patients with hematologic malignancies. Eur J Radiol 2017;93:284–8. 10.1016/j.ejrad.2017.05.046

8. Ohzu M, Takazawa H, Furukawa S, et al. Anorectal abscess in a patient with neutropenia and refractory acute myeloid leukemia: to operate or not to operate Am J Case Rep 2021;22:e931589. 10.12659/AJCR.931589

9. Gu X, Wang L, Fu A, et al. Efficacy of 1.22% metronidazole and sodium chloride hip bath combined hemorrhoid suppository against perianal infection in acute leukemia patients after chemotherap. Chin Hosp Pharm J 2017;37:1281–4.

10. Sun Y, Hu Z, YH M. Influencing factors of rural women’s cervical cancerscreening behavior intention based on the theory of planned behavior. China Cancer 2022;31:534–42.

11. Donders ART, van der Heijden GJMG, Stijnen T, et al. Review: a gentle introduction to imputation of missing values. J Clin Epidemiol 2006;59:1087–91. 10.1016/j.jclinepi.2006.01.014

12. Marquaridt DW. Generalized inverses, ridge regression, biased linear estimation, and nonlinear estimation. Technometrics 1970;12:591–612. 10.1080/00401706.1970.10488699

13. Chen X, Zhu X, Tian M, et al. Construction and validation of a prediction model for pulmonary infection in elderly patients with hip fracture after surgery. Chinese Journal of Nursing 2021;56:659–66.

14. He J, Sun J, Zhang T, et al. Construction and validation of risk prediction model for deep vein thrombosis in acute exacerbations of chronic obstructive pulmonary disease based on serum angiopoietin 2 levels. Ann Palliat Med 2021;10:7905–18. 10.21037/apm-21-1448

15. Zhong J, He Y, Ma J, et al. Development and validation of a nomogram risk prediction model for malignancy in dermatomyositis patients: a retrospective study. PeerJ 2021;9:e12626. 10.7717/peerj.12626

16. Lohsiriwat V. Treatment of hemorrhoids: a coloproctologist’s view. World J Gastroenterol 2015;21:9245–52. 10.3748/wjg.v21.i31.9245

17. Jiang W, Zhang H, Sui N, et al. Epidemiological survey of common anorectal diseases among urban residents in China. Public Health Chin 2016;32:1293–6.

18. Luo Y, Li M, Chenh Q, et al. Case–control study on risk factors of perianal infection in patients with hematological malignancy undergoing chemotherapy. J Nurs (China) 2022;29:7–11.

19. Bryant AL, Deal AM, Battaglini CL, et al. The effects of exercise on patient-reported outcomes and performance-based physical function in adults with acute leukemia undergoing induction therapy. Integr Cancer Ther 2018;17:263–70. 10.1177/1534735417699881

20. Pan C. To compare the clinical efficacy of diosmin tablets and Diyuhuaijiao pills in the treatment of stage I internal hemorrhoids. Beijing: University of Chinese Medicine, 2015.

21. Loureiro RV, Borges VP, Tomé AL, et al. Anorectal complications in patients with haematological malignancies. Eur J Gastroenterol Hepatol 2018;30:722–6. 10.1097/MEG.0000000000001133

22. Liu H, Wang H, Zhou A, et al. Risk factors for perianal infection in the early stage of hematopoietic stem cell transplantation. Chin J Nosocomiol 2021;31:3782–6.

23. Mo X-D, Yan X, Hu W, et al. Perianal infections in the phase before engraftment after allogeneic hematopoietic stem cell transplantations: a study of the incidence, risk factors, and clinical outcomes. Acta Haematol 2018;139:19–27. 10.1159/000481723

24. Yao X, Wang L, Ding T, et al. Clinical effects of sitting bath with different concentrations of sodium chloride solution combined with 0.1% metronidazole on perianal infection of leukemia patients. Chinese Nursing Research 2020;34:1057–60.

25. Lu L, Ge Y, Zhu X, et al. Effect of iodophor diluent perianal bath on the prevention and treatment of perianal infection in patients with malignant hematological malignancies undergoing chemotherapy. Chinese Journal of Nosocomiology 2019;29:3453–6.

26. Zhou Y, Gao H, Hua H, et al. Clinical effectiveness of matrine sitz bath in treating Perianal infection after chemotherapy for acute leukemia. Ann Palliat Med 2020;9:1109–16. 10.21037/apm-20-912

27. On behalf of the Mucositis Study Group of the Multinational Association of Supportive Care in Cancer/International Society for Oral Oncology (MASCC/ISOO), Hong CHL, Gueiros LA, et al. Systematic review of basic oral care for the management of oral mucositis in cancer patients and clinical practice guidelines. Support Care Cancer 2019;27:3949–67. 10.1007/s00520-019-04848-4

28. Iveson T, Boyd KA, Kerr RS, et al. 3-month versus 6-month adjuvant chemotherapy for patients with high-risk stage II and III colorectal cancer: 3-year follow-up of the SCOT non-inferiority RCT. Health Technol Assess 2019;23:1–88. 10.3310/hta23640

29. Silverman MA, Konnikova L, Gerber JS. Impact of antibiotics on necrotizing enterocolitis and antibiotic-associated diarrhea. Gastroenterol Clin North Am 2017;46:61–76. 10.1016/j.gtc.2016.09.010

30. Amato A, Bottini C, De Nardi P, et al. Evaluation and management of perianal abscess and Anal Fistula: SICCR position statement. Tech Coloproctol 2020;24:127–43. 10.1007/s10151-019-02144-1

31. McQuade RM, Stojanovska V, Abalo R, et al. Chemotherapy-induced constipation and diarrhea: pathophysiology, current and emerging treatments. Front Pharmacol 2016;7:414. 10.3389/fphar.2016.00414

32. Shtyrkova SV, Klyasova GA, Ntanishyan KI, et al. [Perianal infection in patients with hemoblastosis: risk factors and possibilities of prevention]. Ter Arkh 2016;88:72–7. 10.17116/terarkh201688772-77

33. Zheng JX, Wu HD, Zhang YL. Risk factors of perianal infection in patients with acute leukemia and preventive nursing effect analysis. Journal of Taishan Medical College 2017;38:706–7.

34. Lo-Coco F, Avvisati G, Vignetti M, et al. Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Engl J Med 2013;369:111–21. 10.1056/NEJMoa1300874

Associated Data

Supplementary Materials

bmjopen-2023-074196supp001.pdf

bmjopen-2023-074196supp002.pdf

Reviewer comments

Author's manuscript

Data Availability Statement

Data are available on reasonable request.