3 D Statistical Learning

Facebook Twitter Youtube Instagram Linkedin
CONTACT US NOW

Intraoperative Crystalloid and Colloidal Fluids

objectives

The client (An Hospital in Germany) wanted to investigate the impact of intraoperative fluid volume on outcomes in geriatric patients undergoing hip fracture surgery.

Data Information

Retrospective single-center study with data from the hospital information systems was available.  The study included patients aged 70 years or older who had sustained a proximal femur fracture. Patients with pathologic, periprosthetic, or peri-implant fractures and those with missing data were excluded. 

 

Variables in the Data Set

  1. Fallnummer – Case Number
  2. Alter – Age
  3. Geschlecht – Gender
  4. Heimbewohner – Nursing Home Resident
  5. Aufnahme_SAP – Admission (SAP)
  6. Aufnahmeuhrzeit_SAP – Admission Time (SAP)
  7. Operation_SAP – Operation (SAP)
  8. OP Beginn_SAP – Operation Start Time (SAP)
  9. Zeit_Schnitt_Naht – Time from Incision to Suture
  10. Aufnahme (D_U)_SAP – Admission (D_U) (SAP)
  11. OP (D_U)_SAP – Operation (D_U) (SAP)
  12. Datedif_Auf_OP_SAP – Date Difference from Admission to Operation (SAP)
  13. OP Dienst – Operation Service
  14. Todesdatum – Date of Death
  15. Tod Tage postOP – Death Days Post-Operation
  16. Frakturtyp – Fracture Type
  17. ISAR – ISAR (Geriatric Medicine Risk Score)
  18. ASA – ASA (American Society of Anesthesiologists) Classification
  19. Gammanagel – Gamma Nail
  20. Duokopf – Dual Head
  21. HTEP – Hip Total Endoprosthesis
  22. Hämoglobin – Hemoglobin
  23. Leukozyten – Leukocytes
  24. Thrombozyten – Platelets
  25. Kreatinin – Creatinine
  26. CRP – C-Reactive Protein
  27. Quick – Prothrombin Time Quick
  28. INR – International Normalized Ratio
  29. aPPT – Activated Partial Thromboplastin Time
  30. Albumin – Albumin
  31. BMI – Body Mass Index
  32. Größe in m – Height in meters
  33. Gewicht in kg – Weight in kilograms
  34. Zementreaktion_Donaldson – Cement Reaction (Donaldson)
  35. Niereninsuffizienz…35 – Renal Insufficiency (abbreviated)
  36. Antikoagulation – Anticoagulation
  37. Herzinfarkt – Myocardial Infarction
  38. CHF – Congestive Heart Failure
  39. pAVK – Peripheral Arterial Disease
  40. CVA/TIA – Stroke/Transient Ischemic Attack
  41. Demenz – Dementia
  42. Chron Lungen – Chronic Lung Disease
  43. Kollageneose – Collagenosis
  44. Ulkusleiden – Peptic Ulcer Disease
  45. mild Leber – Mild Liver Disease
  46. Diabetes mellitus – Diabetes Mellitus
  47. Hemiplegie – Hemiplegia
  48. Niereninsuffizienz…48 – Renal Insufficiency (abbreviated)
  49. D.M. + Organ – Diabetes Mellitus with Organ Involvement
  50. Tumor – Tumor
  51. Leukämie – Leukemia
  52. Lymphom – Lymphoma
  53. schwer. Leber – Severe Liver Disease
  54. met sol Tumor – Metastatic Solid Tumor
  55. AIDS – AIDS
  56. CCI Summe – Charlson Comorbidity Index Sum
  57. Dindo_Klass – Dindo Classification
  58. Ek Gabe /Anzahl – Ectopic Bone Grafts/Number
  59. HWI – Urinary Tract Infection
  60. Pneumonie – Pneumonia
  61. Wundheilungssstörung – Wound Healing Disturbance
  62. Zweiteingriff – Second Intervention
  63. Sonstige_Komplikation – Other Complications
  64. Intraop_Kristalloid/ml – Intraoperative Crystalloid Fluid Volume (ml)
  65. Intraop_Kolloidal – Intraoperative Colloidal Fluid Volume
  66. Intraop_Katecholamine – Intraoperative Administration of Catecholamines
  67. Reanimation_Intraop – Intraoperative Resuscitation

Grouping variable of interest

For this particular analysis and based on the fluids given, we divided patients into high-volume and low-volume groups: Excessive Intraoperative Fluid Administration (Sum of Intraoperative Crystalloid and Colloidal Fluids > 1500 ml vs. <= 1500):

Group A: Sum of Intraoperative Crystalloid and Colloidal Fluids > 1500 ml 
Group B: Sum of Intraoperative Crystalloid and Colloidal Fluids <= 1500

Univariate Analyse

We employ statistical analyses to derive valuable insights from the dataset, employing a range of methods to enhance our understanding. The provided R code generates descriptive statistics and conducts hypothesis tests, facilitating a comprehensive exploration of the dataset.

The analysis primarily focuses on a detailed comparison of specified variables across distinct groups, specifically evaluating the impact of Excessive Intraoperative Fluid Administration (Sum of Intraoperative Crystalloid and Colloidal Fluids > 1500 ml vs. <= 1500). This comparison is instrumental in understanding the distribution and potential differences for each variable in the two groups.

The statistical tests performed are tailored to the nature of the variables under consideration. For continuous variables, we employ Wilcoxon Rank-Sum tests, while for categorical variables, we perform [insert specific statistical tests used]. This multifaceted approach ensures a thorough examination of the dataset, providing valuable insights that support subsequent statistical inference and decision-making processes.

 
 
 

Multivariate Analysis

Multivariate Analysis, adjusted for age, gender, American Society of Anesthesiologists (ASA) grade, CCI , fracture type : 

 

 

For each of the following outcomes, multiple logistic regressions were performed, each adjusted for age, gender, American Society of Anesthesiologists (ASA) grade, CCI [BD], and fracture type [P]. The objective was to analyze the influence of Excessive Intraoperative Fluid Administration (Sum of Intraoperative Crystalloid and Colloidal Fluids > 1500 ml vs. <= 1500) on the following outcomes: a. Mortality [BE=5] b. Complications (Dindo) [BE] c. Intraop Crystalloid [BL] d. Intraop Colloidal [BM] e. ICU [BW]

objectives

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodoconsequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum. Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum.

Propensity Score Matching

Propensity Score Matching: Adjusted for Age, ASA , Gender, CCI

 

The matching criterion used is the nearest neighborhood. It matches control individuals to the treated group and discards controls who are not selected as matches.  In more details,

The link function used in estimating the distance measure is the logit function. In other words, logistic regression propensity scores (including the covariates given)

 

 

Then, we proceed to the Comparison of matched samples  

Propensity Score Matching: Adjusted for Age [A], ASA [R], Gender [C], CCI [BD]
  1. Research Question: Is there a difference between the groups in terms of:
  2. a. Cement reaction in the OR
  3. b. Mortality
  4. c. Administration of catecholamines
  5. d. Intraoperative crystalloid volume
  6. e. Intraoperative colloid volume
  7. f. Intraoperative resuscitation
  8. g. Postoperative intensive care
  9. h. Central venous catheter rate
  10. i. Artery rate
  11. j. Duration of surgery
 
 

Subanalysis

 
Group C: With colloidal substitution [BM > 0] Group D: Pure crystalloid substitution [BL > 0, BM = 0]

Variables

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud.

R Code for the analysis

For each of the analyses, the R code is available under request. WE have comprehensive and replicable r-codes

Results and conclusion

Patients with a higher American Society of Anesthesiologists (ASA) grade and more comorbidities were more likely to receive more than 1500 ml of fluids. We observed significant differences in anesthesiologic management between the two groups, with a higher rate of invasive blood pressure management (IBP) and central venous catheter usage in the high-volume group. High-volume therapy was associated with a higher rate of complications (69.7% vs. 43.6%, p < 0.01), a higher transfusion rate (odds ratio 1.91 [1.26–2.91]), and an increased likelihood of patients being transferred to an intensive care unit (17.1% vs. 6.4%, p = 0.009). These findings were confirmed after adjusting for ASA grade, age, sex, type of fracture, Identification-of-Seniors-At-Risk (ISAR) score, and intraoperative blood loss.

 

Our study suggests that intraoperative fluid volume is a significant factor that impacts the outcome of hip fracture surgery in geriatric patients. High-volume therapy was associated with increased complications.