Risk Factors Analysis and Nomogram Model Establishment of Hidden Blood Loss in Overweight and Obese Elderly Patients After Total Hip Arthroplasty

Figures & data

Table 1 Perioperative Parameters of Patients

parameter	Mean ± SD or n (%)
Hemoglobin loss (g/L)	32.1±12.1
Hematocrit level loss (%)	9.1±3.7
Hemoglobin <100 g/L preoperatively (n)	8 (2.3%)
Hemoglobin <100 g/L postoperatively (n)	158 (44.6%)
Visible blood loss (mL)	193±112
TBL (mL)	1104±468
HBL (mL)	911±438
Percentage of HBL in TBL (%)	79.5%±13.1%
Percentage of HBL in EBV (%)	21.8%±10.5%

Abbreviation: TBL, total blood loss; HBL, hidden blood loss; EBV, estimated blood volume.

Table 2 Results of Correlation Analysis of HBL and Independent Samples t-Test for Different Counting Data Groups. Data are Expressed as Mean ±SD *P<0.05

Index	Groups	n	HBL(ml)	T	P
Sex	Male	146	963±439	1.878	0.061
	Female	208	874±435
Diagnosis*	Femoral neck fracture	139	666±369	9.454	0.000
	Femoral head necrosis	215	1069±406
Hypertension*	With	185	763±401	7.108	0.000
	Without	169	1073±421
Diabetes	With	47	825±445	1.453	0.147
	Without	307	924±437
Cardiovascular disease	With	47	822±457	1.497	0.135
	Without	307	925±435
ASA*	≥3	89	785±434	3.168	0.002
	<3	265	953±433

Abbreviation: ASA, American Society of Anesthesiologists.

Table 3 Results of Various Measurement Data and Pearson Correlation Analysis in the Perioperative Period of Patients. Data are Expressed as Mean ±SD, *P<0.05

parameter	N (%) or mean ± SD	Pearson value	P value
Age (years) *	65.21±10.70	−0.144	0.007
The operation time*	78.85±28.77	0.218	0.000
Preoperative RBC*	4.42±0.54	0.278	0.000
Preoperative PLT	197.81±63.37	−0.064	0.232
Preoperative PT	13.05±0.99	−0.008	0.874
Preoperative APTT	37.64±4.19	0.040	0.455
Preoperative INR	0.99±0.10	0.004	0.938
Preoperative TT	17.49±1.32	−0.019	0.725
Preoperative FIB	3.68±0.87	−0.018	0.737
Preoperative D-Dimer*	2.37±3.69	−0.149	0.005
Preoperative blood calcium*	2.33±0.12	0.245	0.000
Preoperative ESR	21.10±19.25	−0.005	0.930
Intraoperative blood loss*	192.68±111.94	0.146	0.006
Estimated blood volume*	4215.62±593.78	0.170	0.001

Abbreviation: RBC, red blood cell; PLT, platelet; PT, prothrombin time; APTT, activated partial thrombin time; INR, international normalized ratio; TT, thrombin time; FIB, fibrinogen; ESR, erythrocyte sedimentation rate.

Table 4 Multiple Linear Regression Analysis of Risk Factors of HBL After THA. *P<0.05. Multiple Linear Regression Analysis Adjusted for All Variables in the Table. Adjusted R² =0.319

	Coefficient B	SE	T value	P value
Constant	12.404	479.780	0.026	0.979
Age	2.398	2.218	1.081	0.281
Diagnosis*	−344.397	48.395	−7.116	0.000
ASA score	6.610	50.461	0.131	0.896
Hypertension*	−257.271	42.329	−6.078	0.000
The operation time*	2.445	0.758	3.225	0.001
Preoperative RBC*	104.005	41.251	2.521	0.012
Preoperative DD polymer*	11.947	6.015	1.986	0.048
Preoperative blood calcium	131.761	183.135	0.719	0.472
Intraoperative blood loss	0.114	0.196	0.583	0.561

Abbreviation: ASA, American Society of Anesthesiologists; RBC, red blood cell.

Figure 1 The nomogram constructed using HBL = 1000 mL as the predicted outcome in overweight and obese patients after THA. The model involves five factors: diagnosis, hypertension, operative time, preoperative erythrocytes, and preoperative D-dimer levels. The number of points for each of the five indicators on the top dotted axis of the nomogram are summed to obtain the total points, which corresponding to the bottom axis regarding probability that the patient will reach 1000 ml of HBL after THA. For example, Patient A had the following results: a diagnosis of femoral head necrosis, hypertension, preoperative RBC: 4.63, operative time: 75 min, preoperative D-dimer level: 0.51 (all marked with red dots in the graph); the final sum of each score is 221, which corresponds to a 39.8% probability of reaching 1000 ml of HBL postoperatively in this patient.

Figure 2 The ROC curve of HBL>1000ml in overweight and obese patients after THA. (A) The area under the curve was 0.758 and the confidence interval is 0.707 to 0.809 in training group; (B) The area under the curve was 0.736 and the confidence interval is 0.653 to 0.819 in validation group.

Figure 3 The calibration curves for the prediction of HBL>1000ml in overweight and obese patients after THA. (A) the training group; (B) the validation group.

Figure 4 The DCA of the nomogram for the HBL prediction in overweight and obese patients after THA.

Note: (A) HBL>1000ml benefit in the training cohort. (B) HBL>1000ml benefit in the validation cohort.