Fully Threaded Versus Partially Threaded Cannulated Screws in the Treatment of Nondisplaced or Valgus-Impacted Geriatric Femoral Neck Fractures: A Retrospective Cohort Study

Abstract

Introduction:

Partially threaded cannulated screws (PTCS) have been associated with high rates of femoral neck shortening (FNSH) and complications in the treatment of geriatric femoral neck fractures. There are limited studies directly comparing the use of all fully threaded cannulated screws (FTCS) with PTCS. The purpose of this study was to compare outcomes between FTCS versus PTCS.

Methods:

This retrospective cohort study took place at one Level I Trauma Center and one community hospital between 2011 and 2025. Patients aged ≥65 years treated with PTCS (n = 130) or FTCS (n = 34) for femoral neck fractures were included. The primary outcome was FNSH. Secondary outcomes were overall complications, osteonecrosis, nonunion/fixation failure, salvage arthroplasty, and revision surgery. Penalized regression adjusting for covariates and sensitivity analysis restricted to patients with ≥12-month follow-up were performed.

Results:

On penalized regression analysis, FTCS was associated with lower rates of FNSH (P = 0.011), salvage arthroplasty (P = 0.045), and overall complications (P = 0.022) compared with PTCS. On 12-month sensitivity analysis, findings remained consistent for FNSH (P = 0.030) and were also notable for lower nonunion/fixation failure with FTCS (P = 0.026).

Conclusion:

FTCS were associated with markedly lower rates of FNSH across all analyses. FTCS demonstrated lower rates of salvage arthroplasty and overall complications on primary and adjusted analyses. On sensitivity analysis, arthroplasty and complication rates showed similar directional trends but did not reach statistical significance. Compared with PTCS, FTCS did not demonstrate higher rates of osteonecrosis, nonunion/fixation failure, or revision surgery. These findings suggest that FTCS represent a safe and reliable option for this injury pattern.

Ayoub and Martirosian are part owners of Central California Surgical Distributors (CCSD). None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Dr. Saucedo, Robles, and Vue.

The treatment of nondisplaced and valgus-impacted femoral neck fractures in the elderly is controversial.¹ Internal fixation is common in these fractures but is associated with a high failure rate. Although posterior tilt is a risk factor for fixation failure,^2,3 even truly nondisplaced fractures without sagittal malalignment have shown a high revision surgery and complication rate.¹

Multiple methods of fixation have been attempted for these fractures, including the sliding hip screw,⁴ femoral neck system,^5,6 antishortening screws,⁷ headless compression screws,⁸ and hybrid constructs with both partial and fully threaded screws.^9,10 These alternate constructs are associated with complications, including higher rates of osteonecrosis compared with cannulated screw fixation,⁴ femoral neck shortening,^4,7 and difficulty with implant removal.¹¹

One of the most common fixation constructs for femoral neck fractures remains an inverted triangle formation with partially threaded cannulated screws (PTCS).^10,12,13 PTCS allow the fracture ends to compress against one another, which theoretically enhances healing and decrease nonunion rates.¹² Although there is benefit to fracture compression, multiple studies have reported a revision surgery rate with PTCS of more than 20% and nearly 30% rate of femoral neck shortening (FNSH) greater than 5 mm.^4,14 The clinical significance of femoral neck shortening is reflected in lower functional outcomes, likely related to abnormal abductor moments and altered gait.^9,12,15

Multiple studies have sought to address concerns regarding femoral neck shortening (FNSH) and the overall high complication rates associated with PTCS by incorporating fully threaded screws into their constructs. Most have compared hybrid constructs (a mix of partially and fully threaded screws) with PTCS, have used headless compression screws, or have included patients younger than 60 years, in whom bone quality may differ inherently. To the authors' knowledge, no study has directly compared all fully threaded cannulated screws (FTCS) to PTCS in the geriatric population.

The purpose of this study was to directly compare PTCS with FTCS for nondisplaced and valgus-impacted femoral neck fractures in the geriatric population. The authors hypothesized that FTCS would have less FNSH and fewer complications than partially threaded screws for stable femoral neck fractures.

Methods

This was an institutional review board approved retrospective study. All patients with nondisplaced or valgus-impacted femoral neck fractures, OTA/AO classification 31B1, 31B2,¹⁶ were assessed. Femoral neck fractures were treated by a total of eight fellowship-trained orthopaedic trauma surgeons at a Level I Trauma Center and a community hospital between 2011 and 2025. Patients were identified by CPT code 27235. Inclusion criteria were patients ages 65 years or older treated with percutaneous fixation with large fragment cannulated screws (6.5 mm diameter or larger) for a femoral neck fracture. Radiographs and surgical reports for 300 patients were reviewed by two fellowship-trained orthopaedic trauma surgeons and an orthopaedic resident to identify patients meeting inclusion criteria. Exclusion criteria were pathologic fractures, atypical constructs (fewer than three screws or unsatisfactory screw placement), hybrid implants (combining partially and fully threaded screws), death within 3 months, open reductions or capsulotomies, inadequate preoperative imaging, and notable initial displacement.

Notable initial displacement was defined as any varus displacement on AP view or sagittal tilt greater than 20° on the lateral view on preoperative or intraoperative imaging. Preoperative imaging was considered appropriate if there was a frog leg or cross table lateral radiograph, or a CT scan of the hip. Intraoperative screw position was assessed by two fellowship-trained orthopaedic trauma surgeons to confirm they were satisfactory. Factors such as appropriate spread on AP and lateral views, avoidance of in-out-in screws, and avoidance of a start point below the lesser trochanter for all screws were evaluated.

FNSH >5 mm, nonunion, revision surgery, osteonecrosis, presence of complications, washer use, Garden I versus Garden II classification, and postoperative weight-bearing were identified from the medical record and from imaging. FNSH was evaluated using the final follow-up notes and radiographs by a fellowship-trained orthopaedic trauma attending or orthopaedic trauma fellow. Although formal interobserver and intraobserver reliability testing was not done, measurements followed a previously described and widely reported technique used in large multicenter trials.^17-19 Femoral neck shortening was defined as the change in screw head protuberance—from the lateral femoral cortex to the most prominent point of the cannulated head along the screw axis-calibrated—using the known screw length (Figure 1). A threshold of ≥5 mm was used to define shortening, consistent with prior studies.^7,17,20,21 Complications were defined as radiographic evidence of nonunion/fixation failure, osteonecrosis, FNSH > 5 mm, or screw penetration. Given that the fracture line was not always clearly visible radiographically, the diagnosis of nonunion was made using a combination of symptoms, radiographic evolution, and/or fixation failure rather than a single early radiographic timepoint. In addition, a sensitivity analysis restricted to patients with ≥12-month follow-up was used to reduce the risk of misclassification related to limited follow-up duration. Revision surgeries were included only when attributable to complications of the index fracture fixation, including fixation failure, nonunion, osteonecrosis, or other implant-related complications. Procedures performed for unrelated indications were excluded.

Figure 1.

Radiographic image showing measurement technique for postoperative femoral neck shortening after cannulated screw fixation. Femoral neck shortening defined as the distance from the lateral femoral cortex to the start of the cannulated screw head. Calibration done by dividing the known screw length by the measured length.

When medical records and imaging were not available for at least 3 months postoperatively, patients and their families were contacted through phone to obtain outcomes of revision surgery or conversion to arthroplasty. For other outcomes requiring radiographic follow-up (i.e., nonunion, osteonecrosis, FNSH, or complications), a subgroup analysis was done limited only to those patients with at least 3 months of imaging postoperatively. Patients without medical records and imaging past 3 months that were unable to be reached by phone were considered lost to follow-up and excluded from the study.

Patient characteristics and categorical outcomes were compared using chi-square and Fisher exact test as appropriate based on cell counts. To address potential bias from rare events and complete separation, penalized logistic regression using Firth method was done for adjusted analyses. Multivariable covariates were selected a priori based on clinical relevance and potential to confound the association between fixation type and outcomes, rather than on univariate significance testing. These included age, smoking status, and year of surgery. Given the limited number of events, the number of covariates was restricted to avoid model overfitting and unstable estimates.

Penalized logistic regression using Firth method included these covariates and was done for all outcomes to address complete separation and rare events. Firth penalized logistic regression was implemented using the logistf package in R. Statistical analyses were done using SPSS version 31.0 (IBM Corp., Armonk, NY) and R statistical software (R Foundation for Statistical Computing, Vienna, Austria). Sensitivity analysis restricted to patients with ≥12-months of radiographic follow-up was done for nonunion/fixation failure, osteonecrosis, FNSH, and overall complications.

Results

Patients

Out of 300 patients screened, 164 (130 treated with PTCS and 34 with FTCS) met inclusion criteria and were included in the final analysis of revision surgery and salvage arthroplasty. Overall, 15.0% of patients lost to follow-up, defined as less than 3 months postoperatively who were excluded. Although more patients were lost to follow-up in the PTCS group compared with the FTCS group, this did not reach statistical significance (28 patients, 17.7%, versus one patient, 2.9%, respectively, P = 0.17). In addition, PTCS had longer average follow-up than FTCS (25.0 months versus 15.2 months, respectively, P = 0.001). Median clinical follow-up for PTCS versus FTCS was 18.1 months (7 to 30.9 months) and 14.0 months (6.3 to 20.5 months), respectively. Of the total cohort, 148 patients (116 PTCS, 32 FTCS) had radiographic follow-up beyond 3 months and comprised the subgroup analysis evaluating FNSH, nonunion/fixation failure, osteonecrosis, and overall complications. No difference was found in radiographic follow-up between the groups (PTCS 14.6 months, FTCS 12.1 months, P = 0.32). Median radiographic follow-up for the PTCS and FTCS patients were 8 months (3.5 to 18.5 months) and 8.4 months (4.3 to 17.4 months), respectively.

The average age for all patients was 79.3 years (range 65 to 98 years). Overall, 123 female (75.0%) and 41 male (25.0%) patients were there. Patient demographics are outlined in Table 1. The FTCS cohort demonstrated a higher rate of Garden I fractures and lower washer utilization, although these differences did not reach statistical significance (Table 1). No difference was observed in chronic kidney disease, diabetes mellitus, smoking status, or immediate postoperative weight-bearing status between the groups (Table 1).

Table 1.

Geriatric Femoral Neck Fracture Patient Demographics

	Total (n = 164)	Fully Threaded Cannulated Screws (FTCS)	Partially Threaded Cannulated Screws (PTCS)	P
Patients, n (%)	164 (100)	34 (20.7)	130 (79.3)
Sex, n (%)
Female	123 (75.0)	24 (70.6)	99 (76.2)	0.52
Male	41 (25.0)	10 (29.4)	31 (23.8)
Age, mean (range)	79.3 (65-98)	80.9 (65-95)	78.9 (65-98)	0.20
Chronic kidney disease (n)	16.5%	8.8% (3)	18.5% (24)	0.28
Smoking (n)	18.9%	5.9% (2)	19.2% (29)	0.09
Diabetes mellitus (n)	26.2%	32.4% (11)	24.6% (32)	0.36
Washer use	18.3%	5.9% (2)	21.5% (28)	0.06
Garden I	65.9%	79.4% (27)	62.3% (81)	0.07
Garden II	34.1%	20.6% (7)	37.7% (49)
Non weight-bearing postoperatively (n)	13.4%	14.7% (5)	13.1% (17)	1.00
Mean follow-up (SD), months	23.0 (23.7)	15.2 (10.5)	25.0 (25.7)	0.001
Median follow-up [IQR]	16.4 [7-28.3]	14.0 [6.3-20.5]	18.1 [7-30.9]	—
Mean radiographic follow-up (SD), months	14.0 (17.8)	12.1 (10.1)	14.6 (19.4)	0.32
Median radiographic follow-up [IQR]	7.9 [3.9-18.3]	8.4 [4.3-17.4]	8 [3.5-18.5]	—

Bold entries indicate statistial significance (p<0.05).

IQR = interquartile range, SD = standard deviation.

Outcomes Requiring Radiographic Follow-up

In the subgroup analysis of patients with radiographic follow-up beyond 3 months (Table 2), there was a markedly lower incidence of FNSH in the FTCS group compared with the PTCS group in unadjusted analysis (0% versus 19.0%, P = 0.003). Overall complication rates were also lower in FTCS compared with PTCS (3.1% versus 24.1%, P = 0.01). Penalized logistic regression adjusting for covariates confirmed these findings, with adjusted odds ratios and 95% confidence intervals reported in Table 2 (P < 0.05). Sensitivity analysis >12-months demonstrated similar findings for FNSH (P < 0.05) and showed a trend toward decreased overall complications in the FTCS; however, this did not reach statistical significance (P = 0.078). Sensitivity analysis ≥12-months showed a decrease in nonunion/fixation failure in the FTCS (P = 0.026) that was not seen on unadjusted or regression analysis (P > 0.05).

Table 2.

Comparison of Femoral Neck Shortening, Nonunion, Osteonecrosis, and Overall Complications for Partially Threaded Versus Fully Threaded Screws, Subgroup Analysis With Radiographic Follow-up

	Partially Threaded Screws (n = 116)	Fully Threaded Screws (n = 32)	Total (n = 148)	P	Adjusted OR (Firth)	95% CI	P	Adjusted OR (>12 mo)	P (>12 mo)
Femoral neck shortening (n)	19.0% (22)	0% (0) a	14.9% (22)	0.003	0.070	0.0005-0.615	0.011	0.084	0.030
Nonunion/fixation failure (n)	7.76% (9)	0% (0)a	6.1% (9)	0.11	0.212	0.198-1.25	0.089	0.075	0.026
Osteonecrosis (n)	7.8% (9)	3.1% (1)	6.8% (10)	0.68	2.16	0.163-23.13	0.524	0.17	0.175
Overall complications (n)	24.1% (28)	3.1% (1)	19.6% (29)	0.01	0.166	0.0169-0.792	0.022	0.22	0.078

Bold entries indicate statistial significance (p<0.05).

CI = confidence interval, OR = odds ratio.

^aExact binomial 95% confidence interval for zero-event outcomes in the fully threaded screws cohort was 0 to 10.9%.

No notable differences were observed between the groups for osteonecrosis or revision surgery on primary and adjusted analyses. Year of surgery was associated with increased osteonecrosis (P = 0.014) on adjusted analysis; however, this association was not maintained in sensitivity analysis ≥12-months (P = 0.361).

Outcomes Not Requiring Radiographic Follow-up

Primary analysis of all patients (Table 3) demonstrated a lower rate of revision to arthroplasty in the FTCS group compared with PTCS (0% versus 11.5%, P = 0.03). Adjusted analysis using penalized logistic regression demonstrated similar results, with adjusted odds ratios and 95% confidence intervals reported in Table 3 (P = 0.045). Indications for salvage arthroplasty included symptomatic loss of reduction, malunion, or osteonecrosis (Figure 2). The overall revision surgery rate was lower in the FTCS group (2.9% versus 14.6%), although this difference did not reach statistical significance on primary, penalized regression, or 12-month sensitivity analyses. Among the PTCS group, revision surgeries included 15 revisions to arthroplasty (8 total hip arthroplasties, seven hemiarthroplasties), with the remaining cases involving implant removal or screw exchange. In the FTCS group, the single revision surgery was for screw penetration related to osteonecrosis, which was managed with implant removal (Figure 3).

Table 3.

Comparison of Salvage Arthroplasty and Revision Surgery for Partially Threaded Versus Fully Threaded Screws for Geriatric Femoral Neck Fracture Patients

	Partially Threaded Screws (n = 130)	Fully Threaded Screws (n = 34)	Total (n = 164)	P	Adjusted OR (Firth)	95% CI	P
Salvage arthroplasty (n)	11.5% (15)	0% (0)a	9.1% (15)	0.03	0.094	0.0007-0.955	0.045
Revision surgery (n)	14.6% (19)	2.9% (1)	12.2% (20)	0.095	0.32	0.03-1.78	0.209

Bold entries indicate statistial significance (p<0.05).

CI = confidence interval, OR = odds ratio.

^aExact binomial 95% confidence interval for the arthroplasty rate in the fully threaded screws cohort was 0 to 10.3%.

Figure 2.

Images showing valgus-impacted femoral neck fracture treated with partially threaded cannulated screws. A, Intraoperative fluoroscopy showing left femoral neck fracture treated with three partially threaded cannulated screws in an inverted triangle construct. B, Three-week follow-up showing loss of reduction with varus collapse and shortening. C, One year follow-up showing screw penetration on coronal CT (left image) and revision to total hip arthroplasty (right image).

Figure 3.

Images showing valgus-impacted femoral neck fracture treated with fully threaded cannulated screws. A, Immediate postoperative radiographs showing left femoral neck fracture treated with three fully threaded cannulated screws in inverted triangle construct. B, One year follow-up anterior-posterior radiograph showing screw penetration and femoral head collapse consistent with osteonecrosis treated with implant removal.

Covariate analysis showed smoking was associated with increased rates of arthroplasty (P = 0.036) and increased age was associated with lower rates of arthroplasty and revision surgery (P = 0.003, P = 0.022, respectively); however, these results were not maintained in 12-month sensitivity analyses.

Time-To-Event Analysis

The mean time to complications, revision surgery, and arthroplasty was 4.8, 12.5, and 13.6 months, respectively. Median time to each event was 4, 10, and 10 months, respectively (Table 4). Kaplan-Meier analysis (Figure 4) demonstrated no notable differences in arthroplasty-free or revision surgery–free survival between the groups (log-rank P = 0.081 and P = 0.146, respectively), although a trend toward arthroplasty-free survival was observed.

Table 4.

Time-To-Event Outcomes and Follow-up Duration: Geriatric Femoral Neck Fracture Patients

	Mean (SD), months	Median [IQR]
Time to complication	4.8 (3.4)	4 [2-6.3]
Time to revision surgery	12.5 (10.3)	10 [4.4-15.6]
Time to arthroplasty	13.6 (11.4)	10 [4.3-19]
FTCS clinical follow-up	15.2 (10.5)	14.0 [6.3-20.5]
FTCS radiographic follow-up	12.1 (10.1)	8.4 [4.3-17.4]
Total clinical follow-up	23.0 (23.7)	16.4 [7-28.3]
Total radiographic follow-up	14.0 (17.8)	7.9 [3.9-18.3]

FTCS = fully threaded cannulated screws, IQR = interquartile range, SD = standard deviation.

Figure 4.

Kaplan–Meier survival curves comparing fully threaded cannulated screws (FTCS) and partially threaded cannulated screws (PTCS). A, Arthroplasty-free survival. B, Revision surgery–free survival. Tick marks indicate censored observations. Curves are displayed through 50 months of follow-up for clarity. No notable differences were observed between the groups (log-rank P = 0.081 and P = 0.146, respectively).

Discussion

This study found that in the treatment of geriatric femoral neck fractures, FTCS were associated with a lower incidence of FNSH across primary, adjusted, and sensitivity analyses, and a similar trend toward lower salvage arthroplasty and overall complication rates compared with standard PTCS. Primary and adjusted regression analysis showed a statistically significant decrease in salvage arthroplasty and complications in the FTCS group; however, these associations did not remain statistically significant on 12-month sensitivity analysis. Although both unadjusted and adjusted analyses showed no difference in nonunion/fixation failure between the groups, 12-month sensitivity analysis identified a statistically significant association favoring FTCS. No difference was found in revision surgery or osteonecrosis rates between the groups on all analyses. These findings may suggest that in an already impacted, stable fracture, there is no benefit to additional compression through a partially threaded or headless compression screw.

The associations between year of surgery with osteonecrosis, smoking with arthroplasty, and age with arthroplasty/revision surgery observed on adjusted analysis were not maintained in sensitivity analysis, suggesting that these findings may reflect differences in follow-up duration or limited event counts and statistical power rather than true effects. For example, increased age may be associated with earlier mortality and shorter follow-up, accounting for lower observed rates of revision surgery and arthroplasty rather than a true protective effect. Similarly, the lack of statistical significance for some outcomes on ≥12-month sensitivity analysis may reflect limited event counts and insufficient power rather than a true absence of an effect. In addition, baseline differences were found between the groups that approached statistical significance, including a higher rate of Garden I fractures in the FTCS cohort and lower washer utilization. These differences represent potential sources of residual confounding despite adjusted analysis. In addition, loss to follow-up was greater in the PTCS cohort, and outcomes among these patients remain unknown.

Prior biomechanical studies comparing fully and partially threaded screws are mixed: some have suggested more stability when adding a fully threaded screw to the construct,¹⁰ whereas others found more stability with partially threaded screws, particularly in vertical fractures.²²

Clinical studies directly comparing all FTCS to all PTCS are limited and most involve hybrid constructs. The use of one inferior fully threaded screw showed equivalent outcomes compared with PTCS in one study.¹² Furthermore, in a more recent 2025 cohort, a fully threaded inferior screw was associated with decreased varus collapse.²³ The addition of a posterior FTCS has been shown to be associated with decreased FNSH.²⁴ A recent meta-analysis reported less FNSH with fully threaded screws; however, this included headless compression screws, hybrid constructs combining partially and fully threaded screws, and 7 of 11 of their studies focused on a younger population with an average age of 60 years or younger.²⁵

Multiple studies have suggested older age is the primary risk factor for FNSH.^26,27 Chiang et al²⁶ found no difference in complications between fully threaded and partially threaded screws, with age being the only notable risk factor for >5 mm FNSH. DeClouette et al²⁸ also reported that FNSH did not differ based on implant configuration. Wang et al⁸ showed higher screw cutouts in displaced femoral neck fractures with the use of fully threaded headless cannulated screws and concluded that their potential advantages need to be further proven in nondisplaced femoral neck fractures.

Conversely, Weil et al⁹ did report decreased FNSH using fully threaded screws; however, their FTCS group had a 37% loss to follow-up and included hybrid constructs with one partially threaded screw. Another study found no difference in Harris Hip Scores or Radiographic Union Scale for Hip (RUSH) at 6 weeks, and 3 months between partial and fully threaded screws, but improved scores for fully threaded screws at 6 months.¹⁵ They did not report on overall complications or revision surgeries.

This study is consistent with others, suggesting that FNSH can be a notable sequela of using PTCS to treat femoral neck fractures and that this may be mitigated using fully threaded constructs.^9,25 The revision surgery rate for the PTCS group was substantial (14.6%); this was consistent with other reports showing a relatively high rate of revision surgery.^4,14 The study was underpowered to show a difference in revision surgery rate; this could explain why there was no statistically significant difference found between the groups despite a nearly 5 times higher revision surgery rate in the PTCS group. Primary and regression analyses demonstrated a higher rate of revision to arthroplasty for PTCS compared with FTCS, which has not been previously reported. This is a notable finding; although there is some evidence of acceptable outcomes with salvage total hip arthroplasty (THA) for failed cannulated screw fixation,¹³ there are multiple series that have reported higher instability, dislocation, infection, need for revision, and mortality with salvage arthroplasty compared with primary arthroplasty for femoral neck fractures.^29-32 It should also be noted that in the current investigation, fully threaded screws were introduced during the final five years of the study period: primarily for patients with notably poor bone quality to improve screw purchase. Therefore, there may have been a bias toward patients with more osteoporotic bone in the FTCS group.

Previous studies have primarily compared PTCS with hybrid constructs (a combination of partially and fully threaded screws) or headless fully threaded screws and often included younger patient populations.^7-10 This is the first direct comparison of all partially threaded screw construct versus all fully threaded screw construct in the geriatric population that assessed revision surgeries, complications, and femoral neck shortening between the two groups. In addition, the distinction between standard fully threaded screws and headless screws is meaningful, given the notable cost difference between these two implants.³³

There are limitations to this study. First, this is a retrospective comparison with uneven sample sizes. No cases of FNSH and arthroplasty was found in the FTCS group; this is likely related to the small sample size as opposed to truly being a 0% incidence. Second, the inclusion of multiple different surgeons introduces variability and potential confounding to the results. Given the relatively small sample size of the FTCS cohort, inclusion of multiple additional covariates (including individual surgeon identifiers) was not feasible without risking model overfitting; however, the distribution of cases and consistency of findings suggest that surgeon-specific effects are unlikely to account for the observed results. Third, the definition of nonunion at early follow-up may risk misclassification in this population, although sensitivity analysis with longer follow-up was done to address this and yielded similar results. Finally, because FTCS were introduced during the latter portion of the study period, clinical follow-up was shorter in the FTCS group. However, median time to complications, revision surgery, and arthroplasty occurred well within the follow-up period of both groups (Tables 1 and 4), suggesting that these outcomes were unlikely to be missed. Temporal confounding related to evolving surgical techniques, perioperative protocols, and treatment indications over the study period could not be completely excluded despite adjustment for year of surgery. Importantly, patients with sagittal roll back were not included, and this exclusion helped standardize fracture characteristics between the groups.

Conclusion

In this cohort, fully threaded cannulated screws (FTCS) were effective in the treatment of nondisplaced or valgus-impacted geriatric femoral neck fractures. Compared with partially threaded cannulated screws (PTCS), FTCS were associated with markedly lower rates of femoral neck shortening, with a trend toward lower rates of salvage arthroplasty and overall complications, without an increase in osteonecrosis, nonunion, or revision surgery. FTCS appear to be a safe and reliable option for this fracture pattern.