Nomogram to predict postoperative complications in elderly with total hip replacement

Figure 3 Hosmer-Lemeshow test. The null hypothesis is that the fitting probability pi is grouped by 10 decile, and the difference between the fitted value and the observed value in each group, P < 0.05 shows that the scatter separation is significantly deviated from the reference line, the predicted value is not equal to the actual value. Otherwise, the test passes, the predicted value is equal to the actual value.