A 0.05 m Change in Inertial Measurement Unit Placement Alters Time and Frequency Domain Metrics during Running

Inertial measurement units (IMUs) provide exciting opportunities to collect large volumes of running biomechanics data in the real world. IMU signals may, however, be affected by variation in the initial IMU placement or movement of the IMU during use. To quantify the effect that changing an IMU's location has on running data, a reference IMU was 'correctly' placed on the shank, pelvis, or sacrum of 74 participants. A second IMU was 'misplaced' 0.05 m away, simulating a 'worst-case' misplacement or movement. Participants ran over-ground while data were simultaneously recorded from the reference and misplaced IMUs. Differences were captured as root mean square errors (RMSEs) and differences in the absolute peak magnitudes and timings. RMSEs were ≤1 g and ~1 rad/s for all axes and misplacement conditions while mean differences in the peak magnitude and timing reached up to 2.45 g, 2.48 rad/s, and 9.68 ms (depending on the axis and direction of misplacement). To quantify the downstream effects of these differences, initial and terminal contact times and vertical ground reaction forces were derived from both the reference and misplaced IMU. Mean differences reached up to -10.08 ms for contact times and 95.06 N for forces. Finally, the behavior in the frequency domain revealed high coherence between the reference and misplaced IMUs (particularly at frequencies ≤~10 Hz). All differences tended to be exaggerated when data were analyzed using a wearable coordinate system instead of a segment coordinate system. Overall, these results highlight the potential errors that IMU placement and movement can introduce to running biomechanics data.

Acceleration-Based Estimation of Vertical Ground Reaction Forces during Running: A Comparison of Methods across Running Speeds, Surfaces, and Foot Strike Patterns

Twenty-seven methods of estimating vertical ground reaction force first peak, loading rate, second peak, average, and/or time series from a single wearable accelerometer worn on the shank or approximate center of mass during running were compared. Force estimation errors were quantified for 74 participants across different running surfaces, speeds, and foot strike angles and biases, repeatability coefficients, and limits of agreement were modeled with linear mixed effects to quantify the accuracy, reliability, and precision. Several methods accurately and reliably estimated the first peak and loading rate, however, none could do so precisely (the limits of agreement exceeded ±65% of target values). Thus, we do not recommend first peak or loading rate estimation from accelerometers with the methods currently available. In contrast, the second peak, average, and time series could all be estimated accurately, reliably, and precisely with several different methods. Of these, we recommend the 'Pogson' methods due to their accuracy, reliability, and precision as well as their stability across surfaces, speeds, and foot strike angles.

Unsupervised Gait Event Identification with a Single Wearable Accelerometer and/or Gyroscope: A Comparison of Methods across Running Speeds, Surfaces, and Foot Strike Patterns

We evaluated 18 methods capable of identifying initial contact (IC) and terminal contact (TC) gait events during human running using data from a single wearable sensor on the shank or sacrum. We adapted or created code to automatically execute each method, then applied it to identify gait events from 74 runners across different foot strike angles, surfaces, and speeds. To quantify error, estimated gait events were compared to ground truth events from a time-synchronized force plate. Based on our findings, to identify gait events with a wearable on the shank, we recommend the Purcell or Fadillioglu method for IC (biases +17.4 and -24.3 ms; LOAs -96.8 to +131.6 and -137.0 to +88.4 ms) and the Purcell method for TC (bias +3.5 ms; LOAs -143.9 to +150.9 ms). To identify gait events with a wearable on the sacrum, we recommend the Auvinet or Reenalda method for IC (biases -30.4 and +29.0 ms; LOAs -149.2 to +88.5 and -83.3 to +141.3 ms) and the Auvinet method for TC (bias -2.8 ms; LOAs -152.7 to +147.2 ms). Finally, to identify the foot in contact with the ground when using a wearable on the sacrum, we recommend the Lee method (81.9% accuracy).

The medium-term effects of selective dorsal rhizotomy on gait compared to a matched cerebral palsy non-SDR group: A follow-up study

BACKGROUND

Selective dorsal rhizotomy (SDR) has been shown to improve gait in the short-term in children with cerebral palsy (CP). Further study is needed to look at the trajectory of outcomes over the longer-term.

RESEARCH QUESTION

What are the medium-term effects of SDR on gait compared to a matched CP non-SDR group?

METHODS

Participants underwent SDR at mean age 6.3 years and completed baseline, 1-year and 5-year follow-up gait analyses. Non-SDR participants were matched at baseline. Differences were assessed within and between groups. Kinematic variables were analysed using Statistical non-Parametric Mapping (SnPM). Other gait and clinical data were analysed using Friedman's one-way repeated measure analysis of variance and a Mann-Whitney U-test.

RESULTS

The initial SDR group consisted of 29 participants, reducing to 22 at 5-year follow-up. Of these, 15 (68 %) had orthopaedic surgeries either concurrent with or in the intervening period since the SDR, mean 3.3 procedures per participant. The initial non- SDR group had 18 participants, reducing to 17 at 5-year follow-up. Of these, 13 (76 %) had orthopaedic surgeries, mean 5.7 procedures. At 1-year follow-up the SDR group had significantly improved knee extension, ankle dorsiflexion, foot progression, Gait Deviation Index, and normalised step length compared to baseline, p < 0.05, and outcomes were maintained at 5-years. At 1-year follow-up the non-SDR group kinematic patterns were unchanged, but at 5-year follow-up this group demonstrated significantly improved knee extension, ankle dorsiflexion and foot progression. There were no significant kinematic differences between the SDR and the non-SDR group at medium-term follow-up.

SIGNIFICANCE

We have documented the trajectory of gait outcomes post-SDR over 3 assessments and found that short-term gait changes endured in the medium-term. However, kinematic changes were similar to a non-SDR group undergoing routine and orthopaedic care. These outcomes are important to guide surgical decision making and to manage treatment goals and expectations.

Age related progression of clinical measures and gait in ambulant children and youth with bilateral cerebral palsy without a history of surgical intervention

BACKGROUND

Age related progression needs to be considered when assessing current status and treatment outcomes in cerebral palsy (CP).

RESEARCH QUESTION

What is the association between age, gait kinematics and clinical measures in children with bilateral CP?

METHOD

A retrospective database review was conducted. Subjects with bilateral CP with baseline and follow-up 3D gait analyses, but no history of intervening surgery were identified. Clinical and summary kinematic measures were examined for age related change using repeat measures correlation. Interactions with GMFCS classification and whether surgery was recommended were examined using robust linear regression. Timeseries kinematic data for baseline and most recent follow-up analyses were analysed using statistical parametric mapping.

RESULTS

180 subjects were included. 75% of participants were classified as GMFCS I or II at baseline. Mean time to follow-up was 4.89 (2.8) years (range 1-15.9 years) with a mean age of 6.4 (2.4) at baseline and 11.3 (3.4) at final follow-up. 15.5% of subjects demonstrated an improvement in GMFCS classification while GDI remained stable. Age related progression was noted across many clinical measures with moderate correlations (r ≥ 0.5) noted for reduced popliteal angle, long lever hip abduction and internal hip rotation range. In gait, there was reduced hip extension in late stance (p < 0.001), increased knee flexion in mid-stance (p < 0.001), reduced peak knee flexion in swing (p < 0.001) and increased ankle dorsiflexion in stance (p < 0.001). In the coronal plane, there was reduced hip abduction in swing (p < 0.001). In the transverse plane, increased external rotation of the knee (p < 0.001) and reduced external ankle rotation were noted in early stance and through swing (p < 0.001). There were no changes in foot progression or hip rotation.

SIGNIFICANCE

Individuals with CP show age related progression of clinical and kinematic variables. Treatment can only be deemed successful if outcomes exceed or match these age-related changes.

Parental subjective assessment of gait limitations: Comparison with objective gait variables

BACKGROUND

Subjective assessment is an important part of clinical examination providing quality insights into impairments of body structure and functions. Research into the associations between parental perceptions of gait in children with cerebral palsy (CP) and objective clinical gait measures is limited.

RESEARCH QUESTION

What are the parental perceived gait limitations in children with CP and are these perceptions associated with objective clinical gait analysis?

METHODS

Parent questionnaires were retrospectively analysed for children with CP who attended our gait analysis laboratory over a 24-month period. Perceived walking limitations caused by pain, weakness, lack of endurance, mental ability, safety concerns, and balance were recorded on a 5-point Likert scale. Normalised gait speed, normalised step length and the Gait Deviation Index (GDI) were calculated. Differences between responses were assessed using Chi-squared tests with Dunn's post hoc test with Bonferroni adjustment. Spearman's rank correlations were performed to determine the relationship between responses and gait parameters.

RESULTS

Data from 251 participants were included, mean age 9 ± 3.4 years, Gross Motor Function Classification System (GMFCS) level I = 158, II = 64 and III = 29. Balance was perceived to limit walking to the greatest extent, followed by weakness, lack of endurance, safety concerns, pain and mental ability. This rank was consistent across GMFCS levels I, II and III. Perceived balance limitations showed the strongest correlations with objective gait variables, GDI (r = -0.31 p = 0.000), normalised step length (r = -0.30 p = 0.0000) and normalised gait speed (r = -0.24 p = 0.0001).

SIGNIFICANCE

Subjective gait perceptions provide a valuable indication of gait function but are weakly associated with objective clinical gait analysis. Outcome measures that are sensitive to changes in balance may be more responsive to parental concerns and help to satisfy their goals and expectations.

Tibial rotation outcomes following hamstring lengthening as part of single event multilevel surgery in children with cerebral palsy

BACKGROUND

Hamstring lengthening remains the most common surgical procedure in the treatment of crouch gait for children with cerebral palsy (CP). While sagittal plane knee kinematics have been shown to improve post-surgery, the effects on transverse plane kinematics have not been reported. Given the differing actions of the medial and lateral hamstring muscles there is potential for change in tibial rotation post hamstring lengthening.

RESEARCH QUESTION

What is the effect of medial only versus combined medial and lateral hamstring lengthening on tibial rotation during gait in children with CP?

METHODS

A retrospective analysis of children with a diagnosis of CP who underwent a hamstring lengthening procedure. These children were divided into 2 groups: G1 (n = 18) had isolated medial hamstring lengthening while G2 (n = 30) had combined medial and lateral hamstring lengthening. A matched non-surgical control group (n = 15) was also included. Kinematic data were analysed pre and post-operatively. Pre-operative to post-operative outcomes for G1 and G2, a comparison at baseline for both groups and the difference in outcomes between the groups were analysed. Baseline to follow-up outcomes for the control group were also analysed.

RESULTS

External tibial rotation increased significantly within groups (G1: -10°, p < 0.01; G2: -11°, p < 0.001, control: -7.7, p < 0.01), with no significant difference in the change between the intervention groups. Foot progression angles became more external in both intervention groups (G1: -15°, p < 0.001; G2: -15°, p < 0.0001) and did not change in the control group.

SIGNIFICANCE

Results demonstrated similar increases in external tibial rotation, regardless of whether an isolated medial or combined medial and lateral surgery was performed. The control group demonstrated a similar change in external tibial rotation suggesting that hamstring lengthening surgery does not contribute to increased external tibial rotation in children with CP compared to what would be expected due to natural progression.

The influence of crouch gait on sagittal trunk position and lower lumbar spinal loading in children with cerebral palsy

BACKGROUND

Crouch gait is a common pattern in children with CP. Little investigation has been performed as to the role of the trunk during crouch gait. A compensatory movement of the trunk may alter the position of the ground reaction force with the effect of reducing the moment arm about the knee or hip. While this may benefit these joints in the context of reduced loading, there may be implications further up the kinematic chain at the level of the lumbar spine.

RESEARCH QUESTION

Are compensatory movements of the trunk present during crouch gait in children with CP and are levels of loading at the lower lumbar spine affected?

METHODS

A full barefoot lower limb and trunk 3-dimensional kinematic and kinetic analysis, with kinetics estimated at the spinal position of L5/S1, was performed on 3 groups of children, namely CP Crouch, CP No-Crouch and TD. Differences in trunk position and L5/S1 loading were compared between groups.

RESULTS

Mean trunk position in relation to the pelvis and laboratory was not statistically significant between groups. At the level of the spine, no differences were present in mean position between groups for L5/S1 sagittal moment or anterior/posterior force.

SIGNIFICANCE

Crouch gait does not elicit a compensatory response of the trunk in children with CP and, consequently, reactive forces and moments at the lower lumbar spine remain within normal limits. With this in mind, it is unlikely that a crouch gait pattern will affect the health of the spine over time in these children.

Accelerometer-based prediction of running injury in National Collegiate Athletic Association track athletes

Running-related injuries (RRI) may result from accumulated microtrauma caused by combinations of high load magnitudes (vertical ground reaction forces; vGRFs) and numbers (strides). Yet relationships between vGRF and RRI remain unclear - potentially because previous research has largely been constrained to collecting vGRFs in laboratory settings and ignoring relationships between RRI and stride number. In this preliminary proof-of-concept study, we addressed these constraints: Over a 60-day period, each time collegiate athletes (n = 9) ran they wore a hip-mounted activity monitor that collected accelerations throughout the entire run. Accelerations were used to estimate peak vGRF, number of strides, and weighted cumulative loading (sum of peak vGRFs weighted to the 9th power) across the entirety of each run. Runners also reported their post-training pain/fatigue and any RRI that prevented training. Across 419 runs and >2.1 million strides, injured (n = 3) and uninjured (n = 6) participants did not report significantly different pain/fatigue (p = 0.56) or mean number of strides per run (p = 0.91). Injured participants did, however, have significantly greater peak vGRFs (p = 0.01) and weighted cumulative loading per run (p < 0.01). Results from this small but extensively studied sample of elite runners demonstrate that loading profiles (load magnitude-number combinations) quantified with activity monitors can provide valuable information that may prove essential for: (1) testing hypotheses regarding overuse injury mechanisms, (2) developing injury-prediction models, and (3) designing and adjusting athlete- and loading-specific training programs and feedback.

Is adult gait less susceptible than paediatric gait to hip joint centre regression equation error?

Hip joint centre (HJC) regression equation error during paediatric gait has recently been shown to have clinical significance. In relation to adult gait, it has been inferred that comparable errors with children in absolute HJC position may in fact result in less significant kinematic and kinetic error. This study investigated the clinical agreement of three commonly used regression equation sets (Bell et al., Davis et al. and Orthotrak) for adult subjects against the equations of Harrington et al. The relationship between HJC position error and subject size was also investigated for the Davis et al. set. Full 3-dimensional gait analysis was performed on 12 healthy adult subjects with data for each set compared to Harrington et al. The Gait Profile Score, Gait Variable Score and GDI-kinetic were used to assess clinical significance while differences in HJC position between the Davis and Harrington sets were compared to leg length and subject height using regression analysis. A number of statistically significant differences were present in absolute HJC position. However, all sets fell below the clinically significant thresholds (GPS <1.6°, GDI-Kinetic <3.6 points). Linear regression revealed a statistically significant relationship for both increasing leg length and increasing subject height with decreasing error in anterior/posterior and superior/inferior directions. Results confirm a negligible clinical error for adult subjects suggesting that any of the examined sets could be used interchangeably. Decreasing error with both increasing leg length and increasing subject height suggests that the Davis set should be used cautiously on smaller subjects.

Effect of flip-flops on lower limb kinematics during walking: a cross-sectional study using three-dimensional gait analysis

BACKGROUND

Flip-flops are a popular footwear choice in warm weather however their minimalist design offers little support to the foot.

AIM

To investigate the effect of flip-flops on lower limb gait kinematics in healthy adults, to measure adherence between the flip-flop and foot, and to assess the effect on toe clearance in swing.

METHODS

Fifteen healthy adults (8 male, mean age 27 years) completed a three-dimensional gait analysis assessment using Codamotion. Kinematic and lower limb temporal-spatial data were captured using the Modified Helen Hayes marker set with additional markers on the hallux and flip-flop sole.

RESULTS

Compared to barefoot walking, there were no differences in temporal-spatial parameters walking with flip-flops. There was an increase in peak knee flexion in swing (mean difference 4.6°, 95 % confidence interval (CI) [-5.8°, -3.4°], p < 0.001) and peak ankle dorsiflexion at terminal swing (mean difference 2°, 95 % CI [-3°, -1°], p = 0.001). Other kinematic parameters were unchanged. Peak separation between foot and flip-flop was 8.8 cm (SD 1.48), occurring at pre-swing. Minimum toe clearance of the hallux in barefoot walking measured 4.2 cm (SD 0.8). Minimum clearance of the flip-flop was 1.6 cm (SD 0.56).

CONCLUSIONS

Healthy adults adapted well to flip-flops. However, separation of the flip-flop from the foot led to increased knee flexion and ankle dorsiflexion in swing, probably to ensure that the flip-flop did not contact the ground and to maximise adherence to the foot. Minimum clearance of the flip-flop was low compared to barefoot clearance. This may increase the risk of tripping over uneven ground.

Corrections in saccade endpoints scale to the amplitude of target displacements in a double-step paradigm

It is widely held that discrete goal-directed eye movements (saccades) are ballistic in nature because their durations are too short to allow for sensory-based online correction. Recent studies, however, have provided evidence that saccadic endpoints can be mediated via online corrections. Specifically, it has been reported that saccade trajectories adapt to the eccentricity of an unexpectedly perturbed target location (i.e., target 'jump' paradigm). If saccades are subject to online correction mechanisms, then the magnitude of such changes should scale to the amplitude of the target jump. To test this hypothesis, saccadic endpoints for trials on which the target jumped one of three amplitudes (Small: 2.5°, Medium: 5.0°, and Large: 7.5°; i.e., Jump trials) immediately after saccade onset were compared with the endpoints of trials in which the target location did not change (i.e., Reference trials). Results showed that primary saccade endpoints for Jump trials were longer than for Reference trials. Importantly, the magnitude of this increase in endpoint scaled with the amplitude of the target jump. Thus, these results support emerging and coalescent evidence that saccade trajectories are subject to online corrections.

A quantitative comparison of two kinematic protocols for lumbar segment motion during gait

During gait analysis, motion of the lumbar region is tracked either by means of a 2-dimensional assessment with markers placed along the spine or a 3-dimensional assessment treating the lumbar region as a rigid segment. The rigid segment assumption is necessary for inverse dynamic calculations further up the kinematic chain. In the absence of a reference standard, the choice of model is mostly based on clinical experience. However, the potential exists for large differences in kinematic output if different protocols are used. The aim of this study was to determine the influence of using two 3-dimensional lumbar segment protocols on the resultant kinematic output during gait. The first protocol was a skin surface rigid protocol with markers placed across the lumbar region while the second consisted of a rigid cluster utilizing active markers applied over the 3rd lumbar vertebra. Data from both protocols were compared through simultaneous recording during gait. Overall variability was lower in 4 out of 6 measures for the skin surface protocol. Ensemble average graphs demonstrated similar mean profiles between protocols. However, Functional Limits of Agreement demonstrated only a poor to moderate agreement. This trend was confirmed with a poor to moderate waveform similarity (CMC range 0.29-0.71). This study demonstrates that the protocol used to track lumbar segment kinematics is an important consideration for clinical and research purposes. Greater variability recorded by the rigid cluster during lumbar rotation suggests the skin surface protocol may be more suited to studies where axial rotation is a consideration.

The clinical impact of hip joint centre regression equation error on kinematics and kinetics during paediatric gait

Regression equations based on pelvic anatomy are routinely used to estimate the hip joint centre during gait analysis. While the associated errors have been well documented, the clinical significance of these errors has not been reported. This study investigated the clinical agreement of three commonly used regression equation sets (Bell et al., Davis et al. and Orthotrak software) against the equations of Harrington et al. Full 3-dimensional gait analysis was performed on 18 healthy paediatric subjects. Kinematic and kinetic data were calculated using each set of regression equations and compared to Harrington et al. In addition, the Gait Profile Score and GDI-Kinetic were used to assess clinical significance. Bell et al. was the best performing set with differences in Gait Profile Score (0.13°) and GDI-Kinetic (0.84 points) falling below the clinical significance threshold. Small deviations were present for the Orthotrak set for hip abduction moment (0.1 Nm/kg), however differences in Gait Profile Score (0.27°) and GDI-Kinetic (2.26 points) remained below the clinical threshold. Davis et al. showed least agreement with a clinically significant difference in GDI-Kinetic score (4.36 points). It is proposed that Harrington et al. or Bell et al. regression equation sets are used during gait analysis especially where inverse dynamic data are calculated. Orthotrak is a clinically acceptable alternative however clinicians must be aware of the effects of error on hip abduction moment. The Davis et al. set should be used with caution for inverse dynamic analysis as error could be considered clinically meaningful.

How one breaks Fitts's Law and gets away with it: Moving further and faster involves more efficient online control

Adam, Mol, Pratt, and Fischer (2006) reported what they termed "a violation of Fitts's Law" - when participants aimed to targets in an array, movement times (MTs) to the last target location (highest index of difficulty (ID)) were shorter than predicted by Fitts's Law. Based on the results of subsequent studies in which placeholders were present either during planning and/or execution stages of the movements, it was suggested that the violation may emerge because of context-dependent changes in planning processes. The present study examined this planning explanation by conducting detailed kinematic analyses of movements. Participants performed aiming movements to sets of 3 targets in different placeholder arrays with different movement amplitudes. Consistent with previous Fitts's Law violation findings, MTs were not significantly longer for movements to the last versus middle target location. Interestingly, the pattern of peak limb velocities (typically associated with planning processes) did not mirror the changes in MTs. On the other hand, analyses of the effector's spatial variability during the movement suggested greater involvement of online control processes when the target was in the last position. Based on these results, we suggest that the Fitts' Law violation observed here occurred because of more efficient online control processes.

The influence of estimated body segment parameters on predicted joint kinetics during diplegic cerebral palsy gait

Inverse Dynamic calculations are routinely used in joint moment and power estimates during gait with anthropometric data often taken from published sources. Many biomechanical analyses have highlighted the need to obtain subject-specific anthropometric data (e.g. Mass, Centre of Mass, Moments of Inertia) yet the types of imaging techniques required to achieve this are not always available in the clinical setting. Differences in anthropometric sets have been shown to affect the reactive force and moment calculations in normal subjects but the effect on a paediatric diplegic cerebral palsy group has not been investigated. The aim of this study was to investigate the effect of using different anthropometric sets on predicted sagittal plane moments during normal and diplegic cerebral palsy gait. Three published anthropometric sets were applied to the reactive force and moment calculations of 14 Cerebral Palsy and 14 Control subjects. Statistically significant differences were found when comparing the different anthropometric sets but variability in the resulting sagittal plane moment calculations between sets was low (0.01-0.07 Nm/kg). In addition, the GDI-Kinetic, used as an outcome variable to assess whether differences were clinically meaningful, indicated no clinically meaningful difference between sets. The results suggest that the effects of using different anthropometric sets on the kinetic profiles of normal and diplegic cerebral palsy subjects are clinically insignificant.

The difficulty identifying intoeing gait in cerebral palsy

In-toeing in children with cerebral palsy can lead to functional difficulties during gait. This may require surgical management to restore a normal foot progression angle. For this reason it is important to indentify the presence of internal rotation and to establish where the abnormal rotation is occurring. This can be done relatively easily in otherwise healthy subjects by examining foot progression angle as the subject walks towards the assessor. In cerebral palsy the often severely affected gait pattern and potential asymmetry at the pelvis means that in-toeing may be more difficult to identify. Gait laboratory data of 245 subjects with cerebral palsy were studied retrospectively. Of these 102 (41.63%) demonstrated in-toeing relative to the pelvis of one or both limbs. Eleven diplegic subjects (16.18%) in-toed bilaterally giving a total of 113 in-toeing limbs for analysis. Of those, 17 (50%) hemiplegic limbs and 20 (25.32%) diplegic limbs demonstrated a foot progression angle within normal limits due to compensations at the pelvis.

False-positive Mammographic Screening: Factors Influencing Re-attendance over a Decade of Screening

Introduction

International studies on the effect of false-positive mammographic screening results on subsequent re-attendance at screening are inconsistent.

Setting Breast

Check, the national breast screening programme for the Republic of Ireland, screens women two-yearly. Re-attendance for subsequent screening is approximately 90%.

Objectives and Methods

The aim of this research was to quantify the impact of false-positive mammographic screening results on subsequent re-attendance, using the BreastCheck clinical database with a decade of screening and to determine if age group, assessment procedure, initial or subsequent screening, location of appointment for next screening round and time from recall to non-malignant diagnosis predicted re-attendance.

Results

From programme commencement in 2000 to the end of 2007, 13,352 screening tests resulted in assessment; 11,765 participants were aged 50-62 years and of these 9746 received false-positive results (positive predictive value 17.2%). Following a false-positive recall to assessment, re-attendance at subsequent screening differed significantly by procedure type (open biopsy 80.3%; core biopsy only 90.2%; no tissue sampling 91.4%; P < 0.0001). Re-attendance differed significantly by timing of false-positive assessment in a woman's screening history (first versus subsequent screening, 89.5% versus 93.5%, P < 0.0001) and by location of next screening appointment (screening centre 89.8% versus mobile unit 91.3%, P < 0.01). The longer the period between recall to assessment and non-malignant diagnosis the less likely women were to re-attend. After logistic regression, first screening, older age, open surgical biopsy, re-invitation to screening centre and a longer period between recall and non-malignant diagnosis were significant negative predictors of re-attendance.

Conclusion

Since April 2008 BreastCheck has employed full field digital mammography throughout the programme, with a resulting increase in recalls; re-attendance will be closely monitored.