2022 exceptional surveillance of hip fracture: management (NICE guideline CG124)

Cemented vs uncemented

Three sets of comparisons for cemented versus uncemented implants were reported. For THA, there was evidence of no difference in activities of daily living, functional status, HRQoL, mortality, unplanned return to theatre, pain, or adverse events between cemented and uncemented implants.

For HA, there was evidence of no difference in activities of daily living, delirium, unplanned return to theatre or adverse events between cemented and uncemented implants, although it was noted that fewer people had a pulmonary embolism when the HA was fixed without cement. Some evidence showed improved functional status, improved HRQoL, and improved mobility with cemented HA. For mortality at 12 months, moderate-certainty evidence showed that the risk of death at 12 months was reduced using cemented HA (RR 0.86, 95%CI 0.78 to 0.96, 15 studies, 3727 participants). There was evidence of no difference in mortality by fixation technique at 5 years.

For mixed THA and HA comparisons, where participants were randomised to a cemented or uncemented prosthesis but the selection of THA or HA was decided by the treating surgeon and patient, there was evidence of no difference in functional status, HRQoL at 24 months, 12 month or late mortality, or unplanned return to theatre between cemented and uncemented implants. There was evidence of no difference in adverse events but there were fewer interoperative periprosthetic fractures when cement was used.

New evidence on cemented vs. uncemented implants supports the current recommendation on using cemented implants for patients undergoing surgery with arthroplasty.

THA vs HA

There was evidence of no difference according to type of arthroplasty for the following outcomes: activities of daily living, delirium, unplanned return to theatre, discharge destination, HRQoL at 4 months or 9 years post-surgery, or mobility at 3 months, 12 months or 13 years after. There was evidence that HRQoL at 12 months was improved when a THA was used, but this was not likely to be a clinically important difference. For mortality, there was evidence of no difference in mortality by type of arthroplasty at 4 months, 12 months, or late follow-up. For functional status, there was evidence of slight improved functional status within 4 months of surgery, at 12 months, and at more than 24 months for people who received a THA, but none of these effects appeared to suggest a clinically important improvement. There was evidence of no difference in adverse events, although fewer participants had a blood transfusion when a HA was used, however it was noted that this analysis was from only two small studies.

New evidence on THA vs HA supports the recommendation on offering THA rather than HA to patients with a displaced intracapsular hip fracture who meet specific criteria.

Comparisons not directly related to recommendations

The Cochrane review also reported comparisons for single vs multiple articulations of THA, short stem vs standard stem THA and bipolar vs unipolar HA. For all three comparisons, there is little evidence to suggest that one intervention was favoured over the other. One small single study reported fewer intraoperative periprosthetic fractures were noted when a short stem was used. Additionally, one study showed that mobility at 12 months was better when a unipolar HA was used.

Extramedullary implants vs intramedullary nails

There was little evidence of difference between the two interventions for the outcomes of activities of daily living, delirium, functional status at 4 and 12 months post-surgery, HRQoL at 12 months post-surgery, mortality at 4 and 12 months post-surgery, unplanned return to theatre and pain at 4 and 12 months post-surgery. Cephalomedullary nails were favoured over extramedullary implants for independent mobility within 4 months of surgery (RR 1.12, 95% CI 1.01 to 1.23; 7 studies, 719 participants) and a 10-metre walking speed test within 4 months of surgery (MD 0.70, 95% CI 0.63 to 0.77; 1 study, 80 participants). However, this was based on very low certainty evidence and data from a single study respectively.

Cephalomedullary nails were favoured over extramedullary implants for independent mobility after 12 months of surgery when using the Parker 1993 mobility scale (MD 0.48, 95% CI 0.10 to 0.87; 14 studies, 1746 participants; I2 = 63%). There was little evidence of difference between the two interventions for the proportion of people who had independent mobility, the proportion of people who failed to regain their pre-fracture mobility, had sufficient ambulation to perform a Timed Up and Go (TUG) test, the reported time to complete a TUG test and the number of participants who remained in bed, or in a wheelchair.

The Cochrane review reported a significant increase in risk of intraoperative periprosthetic fractures when cephalomedullary nails were used compared to extramedullary implants (RR 2.94, 95% CI 1.65 to 5.24; 35 studies, 4872 participants), as well as a significant increase in risk of postoperative periprosthetic fractures (RR 3.62, 95% CI 2.07 to 6.33; 46 studies, 7021 participants). However, cephalomedullary nails were favoured over extramedullary implants for the outcomes non-union of fracture (RR 0.55, 95% CI 0.32 to 0.96; 40 studies, 4959 participants).

New evidence on extramedullary implants vs intramedullary nails generally supports the recommendation of using extramedullary implants over intramedullary nails in patients with trochanteric fractures. Despite findings indicating that cephalomedullary nails showed improved mobility and walking speed, the certainty of evidence was low and very low making it unlikely to impact existing recommendations. Furthermore, the increased risk of non-union of fracture associated with cephalomedullary nails matched findings reported in the original development of the Hip Fracture: management guideline (NICE guideline CG124).

THA and HA vs other interventions

There was evidence of no clinically important difference between THA and HA compared to other interventions for any reported outcome. From the network meta-analysis, the SUCRA rankings suggested that external fixation and HA may have the greatest likelihood of being ranked higher for unplanned return to theatre, however there are significant overlaps of the 95% confidence intervals.

Despite the lack of clinical important differences between interventions, the possibility of higher ranking of HA for unplanned return to theatre aligns with the recommendation to offer replacement arthroplasty (total hip replacement or hemiarthroplasty) to patients with a displaced intracapsular hip fracture.

THA vs HA

There was evidence of no clinical important difference between THA compared to HA for any outcome. The SUCRA rankings from the network meta-analysis indicated that external fixation and THA may have the greatest likelihood of being ranked higher, and long cephalomedullary nails and HA may have the lowest probability of reducing late mortality (reported as mortality at 12 months). However, the authors note that the 95% CIs of these comparisons of outcomes overlap significantly and so advise caution in drawing meaningful interpretations from the ranking of treatments in this network meta-analysis.

Despite lack of meaningful differences between interventions, the higher ranking of THA and lower ranking of HA for late mortality aligns with the recommendation to offer total hip replacement rather than hemiarthroplasty to patients with a displaced intracapsular hip fracture who meet specific criteria.

Extramedullary implants vs intramedullary nails

In the direct comparisons, there was evidence of a difference in early mortality between dynamic fixed angle plate versus long cephalomedullary nail (RR 1.80, 95% CI 1.02 to 3.18, favours fixed angle plate; 2 studies, 400 participants). There was evidence of no difference between any of the other treatments for this outcome. In the network meta-analysis, there was evidence of no difference between any of the treatments. Although the SUCRA rankings indicated that external fixation and static fixed-angle plate may have the greatest likelihood of being ranked higher, and long and short cephalomedullary nails the lowest probability of reducing early mortality, the authors note that the 95% CIs of these comparisons of outcomes overlap significantly and so advise caution in drawing meaningful interpretations from the ranking of treatments in this network meta-analysis.

The authors indicate, it was not possible to conduct a network meta-analysis for HRQoL. In the direct comparisons, HRQoL at 12 months was improved when a dynamic fixed angle plate was used compared to short cephalomedullary nails, although when the estimate was converted, the difference between treatments was compatible with both no clinically important difference and plausible benefits (MD 3.68, 95% CI 0.94 to 6.42).

In the direct comparisons, there was evidence of a difference in unplanned return to theatre between dynamic fixed angle plate versus condylocephalic nail (RR 3.57, 95% CI 1.91 to 6.66, favour dynamic fixed angle plate; 7 studies, 996 participants). Further network meta-analysis supported these findings (RR 3.33, 95% CI 1.95 to 5.68, favours dynamic fixed angle plate; direct and indirect evidence). There was evidence of no difference between any other treatments for this outcome and no evidence to suggest any one treatment was either substantially better or worse than the other, although it was noted that the estimates had wide 95% CIs indicating substantial imprecision. Network meta-analysis also revealed a difference in unplanned return to theatre between static fixed angle plate versus short cephalomedullary nail (RR 0.45, 95% CI 0.23 to 0.88, favours cephalomedullary nail; direct and indirect evidence).

Condylocephalic nail and static fixed angle plate appeared to have the lowest probability of reducing unplanned return to theatre. Effect estimates for these 2 treatments showed clinically important and statistically significant harms of both treatments compared with the comparator intervention. There was also evidence of the possibility of very substantial harms but also clinically important benefits with long cephalomedullary nails.

New evidence on extramedullary nails vs intramedullary implants supports the recommendation to use extramedullary implants such as a sliding hip screw in preference to an intramedullary nail in patients with trochanteric fractures above and including the lesser trochanter.

Comparisons not directly related to existing recommendations

Network meta-analysis reported the following comparisons for the outcome unplanned return to theatre. Dynamic versus static fixed angle plate (RR 2.48, 95% CI 1.36 to 4.50, favours dynamic fixed angle plate; direct and indirect evidence); static fixed angle plate versus external fixation (RR 0.04, 95% CI 0.00 to 0.80, favours external fixation; indirect evidence); short cephalomedullary nail versus condylocephalic nail (RR 2.98, 95% CI 1.59 to 5.60, favours cephalomedullary nail; indirect evidence) and condylocephalic nail versus external fixation (RR 0.03, 95% CI 0.00 to 0.59, favours external fixation; indirect evidence).

THA and HA vs other interventions

THA with single articulation and cemented and uncemented modern bipolar HA had the greatest likelihood of being ranked higher and therefore improving HRQoL at 12 months, although the certainty of evidence was very low. Arthroplasty treatments had greater likelihood of reducing unplanned return to theatre than internal fixation and non-operative interventions. THAs were also found to rank highest or joint highest out of all interventions for HRQoL at 4 months, 12 months and 24 months. There was evidence of no statistical difference between interventions for early mortality, but THA with single articulation and pins were ranked higher, whilst uncemented bipolar HA was ranked lowest in the network meta-analysis. Network meta-analysis revealed that dynamic fixed angle plates (RR 1.59, 95% CI 1.08 to 2.34; indirect evidence), pins (RR 0.64, 95% CI 0.43 to 0.95; direct and indirect evidence) and screws (RR 0.68, 95% CI 0.48 to 0.95; direct and indirect evidence) were all favoured over THA with single articulation for mortality at 12 months.

The higher ranking of THA with single articulation for HRQoL at 4 months aligns with the recommendation to offer replacement arthroplasty (total hip replacement or hemiarthroplasty) to patients with a displaced intracapsular hip fracture. However, the lower ranking for uncemented bipolar arthroplasty does not align with the recommendation. Furthermore, for mortality at 12 months, multiple interventions were favoured over THA with single articulation which does not support the current recommendation. However, there is uncertainty around the ranking from the NMA due to overlapping of 95% confidence intervals. Additionally, during the development of the guideline the committee determined mortality to be less important than other outcomes, as the interventions were not anticipated to have a significant impact on mortality.

Indirect evidence indicated that dual mobility THA showed clinically important improvements over screw treatment (SMD -1.57, 95% CI -2.62 to -0.53,) and non-operative treatment (SMD -1.65, 95% CI -2.75 to -0.55,) for early HRQoL. Additionally, cemented modern bipolar HA showed clinically important improvements over screw treatment (SMD -0.38, 95% CI -0.64 to -0.13; direct and indirect evidence) and non-operative treatment (SMD -0.46, 95% CI -0.89 to -0.04; indirect evidence) for early HRQoL. Similarly, THA with single articulation was also favoured over screw treatment (SMD -0.43, 95% CI -0.78 to -0.08; indirect evidence) and non-operative treatment (SMD -0.51, 95% CI -1.00 to -0.02, favours THA; indirect evidence) for early HRQoL.

Indirect evidence indicated that both cemented modern bipolar HA and THA with single articulation showed clinically important improvements over dynamic fixed angle plates for HRQoL at 12 months (SMD 0.56, 95% CI 0.08 to 1.05) and (SMD 0.59, 95% CI 0.11 to 1.07) respectively. Additionally, indirect evidence showed uncemented modern unipolar HA demonstrated a clinically important improvement over screw treatment (SMD1.19, 95% CI 0.11 to 2.27) and non-operative treatment (SMD 1.15, 95% CI 0.03 to 2.27).

New evidence on HA and THA vs other interventions supports the recommendation to offer replacement arthroplasty (total hip replacement or hemiarthroplasty) to patients with a displaced intracapsular hip fracture.

Direct comparisons indicated that screw treatment resulted in more unplanned returns to theatre than cemented modern unipolar HA (RR 4.01, 95% CI 1.92 to 8.39; 3 studies, 310 participants), cemented modern bipolar HA (RR 4.35, 95% CI 2.67 to 7.07; 4 studies, 553 participants), uncemented first-generation unipolar HA (RR 5.85, 95% CI 3.47 to 9.87; 2 studies, 515 participants) and THA with single articulation (RR 3.11, 95% CI 2.23 to 4.35; 5 studies, 718 participants). Further network meta-analysis supported these findings.

Additionally, direct evidence showed both cemented modern unipolar HA (RR 10.66, 95% CI 3.85 to 29.50; 2 studies, 233 participants) and cemented modern bipolar HA (RR 0.32, 95% CI 0.15 to 0.65; 2 studies, 226 participants) were favoured over dynamic fixed angle plate for unplanned return to theatre. Following network meta-analysis, these findings were further supported.

Network meta-analysis also indicated that uncemented modern bipolar HA (RR 0.41, 95% CI 0.17 to 0.99; indirect evidence), uncemented first-generation unipolar HA (RR 0.31, 95% CI 0.20 to 0.48; direct and indirect evidence) and THA with single articulation (RR 0.31, 95% CI 0.22 to 0.44, indirect evidence) were favoured over dynamic fixed angle plate for unplanned return to theatre. Furthermore, uncemented modern bipolar HA (RR 2.62, 95% CI 1.11 to 6.16; direct and indirect evidence) was favoured over screw treatment for unplanned return to theatre.

Network meta-analysis also showed more unplanned returns to theatre occurred when pins were used compared to cemented modern unipolar HA (RR 4.16, 95% CI 2.53 to 6.84; indirect evidence), cemented modern bipolar HA (RR 2.96, 95% CI 1.95 to 4.50; indirect evidence), uncemented first-generation unipolar HA (RR 2.91, 95% CI 1.80 to 4.72; indirect evidence) and THA with single articulation (RR 2.86, 95% CI 1.93 to 4.26; direct and indirect evidence). Finally, indirect evidence showed more unplanned returns to theatre when non-operative therapy was compared to the same interventions: (RR 5.41, 95% CI 1.80 to 16.26, favours cemented modern unipolar HA); (RR 3.85, 95% CI 1.35 to 10.99, favours cemented modern bipolar HA); (RR 3.79, 95% CI 1.26 to 11.36, favours uncemented first-generation unipolar HA); (RR 3.73, 95% CI 1.29 to 10.74, favours THA).

New evidence on HA and THA vs other interventions further supports the recommendation to offer replacement arthroplasty (total hip replacement or hemiarthroplasty) to patients with a displaced intracapsular hip fracture.

THA vs HA

Indirect evidence indicated that dual-mobility THA showed clinically important improvements over cemented modern unipolar HA (SMD 1.24, 95% CI 0.21 to 2.28,), uncemented modern bipolar HA (SMD 1.30, 95% CI 0.24 to 2.36,), cemented modern bipolar HA (SMD 1.19, 95% CI 0.16 to 2.21,) and uncemented modern unipolar HA (SMD 1.72, 95% CI 0.61 to 2.82,) for early HRQoL. Furthermore, indirect evidence showed THA with single articulation showed clinically important improvement over cemented modern bipolar HA (SMD 0.66, 95% CI 0.05 to 1.28,) and uncemented modern unipolar HA (SMD 1.48, 95% CI 0.30 to 2.66,) for late HRQoL. There was evidence of no difference between any type of THA and any type of HA with regards to unplanned return to theatre. Direct evidence showed that cemented modern bipolar HA was favoured over THA with single articulation for mortality at 12 months (RR 1.72, 95% CI 1.06 to 2.78; 3 studies, 699 participants).

Network meta-analysis indicated that uncemented first-generation unipolar HA (RR 1.53, 95% CI 1.04 to 2.25; indirect evidence), cemented modern unipolar HA (RR 1.62, 95% CI 1.13 to 2.32, direct and indirect evidence), and cemented modern bipolar HA (RR 1.42, 95% CI 1.01 to 2.00, direct and indirect evidence) were all favoured over THA with single articulation for mortality at 12 months.

Direct comparisons showed cemented modern bipolar HA was favoured over THA with single articulation for late mortality (RR 1.36, 95% CI 1.09 to 1.70; 2 studies, 401 participants).

New evidence on THA vs HA for HRQoL supports the recommendation to offer total hip replacement rather than hemiarthroplasty to patients with a displaced intracapsular hip fracture who meet specific criteria. Alternatively, new evidence for mortality outcomes do not support the recommendation. However, during the development of the guideline the committee determined mortality to be less important than other outcomes, as the interventions were not anticipated to have a significant impact on mortality.

Cemented HA vs Uncemented HA

There was evidence of no difference between interventions for mortality at 4 months, however uncemented first-generation bipolar HA had the worst mean rank and lowest SUCRA values, suggesting this treatment had the lowest probability of reducing early mortality.

Direct and indirect evidence indicated cemented modern bipolar HA was favoured over uncemented modern bipolar HA for mortality at 12 months (RR 0.78, 95% CI 0.61 to 1.00 and RR 1.37, 95% CI 1.02 to 1.85 respectively; 2 studies, 557 participants).

Indirect evidence indicated that cemented modern bipolar HA showed a clinically important improvement in HRQoL at 4 months compared to uncemented modern unipolar HA (SMD -0.53, 95% CI -1.01 to -0.05).

New evidence on cemented HA vs uncemented HA supports the recommendation to use cemented implants in patients undergoing surgery with arthroplasty.

Comparisons not directly related to recommendations

A direct comparison between dynamic fixed angle plate versus screw treatment indicated that screws were favoured for the outcome of late mortality (RR 0.39, 95% CI 0.17 to 0.91; 1 study, 73 participants; direct evidence). Furthermore, cemented modern bipolar HA was favoured over cemented modern unipolar HA for the same outcome (RR 0.79, 95% CI 0.65 to 0.95; direct and indirect evidence).

Direct evidence showed dual-mobility THA was favoured over THA with single articulation for early HRQoL (SMD 1.14, 95% CI 0.10 to 2.17; 1 study, 21 participants). These findings were further supported by network meta-analysis.

Uncemented modern bipolar HA was favoured over uncemented modern unipolar HA for both HRQoL at 12 months (SMD -1.43, 95% CI -2.33 to -0.53; 1 study, 28 participants; direct evidence) and late HRQoL (SMD -1.16, 95% CI -2.15 to -1.17; direct and indirect evidence).