Osteoarthritis is the most common form of arthritis in adults; an estimated 595 million people were living with it in 2020, and the knee is the joint most often affected.1 Its burden is rising in Pakistan, where the age-standardized prevalence increased between 1990 and 2021,2 and a recent community-based study from Nawabshah reported a prevalence of 18.1%.3 Intra-articular injection is part of the guideline-based management of knee osteoarthritis4 and is widely used for osteoarthritis, rheumatoid arthritis and other inflammatory knee conditions.5,6

The benefit of an injection depends on delivering the drug into the joint cavity. Landmark-guided injection is frequently inaccurate: in a fluoroscopically controlled series, the lateral midpatellar portal was intra-articular in 93% of injections, but the anteromedial and anterolateral portals were accurate in only 75% and 71%,7 and a systematic review found that about one in five blind knee injections missed the joint.8 Image guidance improves accuracy across joints,9,10 and ultrasound-guided knee injections have been more accurate than landmark-guided injections in pooled and individual studies (95.8% vs 77.8% and 95.6% vs 77.3%, respectively).11,12 Randomized trials have also shown less procedural pain and better short-term outcomes with ultrasound guidance than with palpation.13–15

Fluoroscopy has traditionally been used to guide joint injections because it shows bony landmarks and allows intra-articular placement to be confirmed with contrast, but it requires a fluoroscopy suite and exposes patients and staff to ionizing radiation.16,17 Ultrasound shows soft tissue, effusion and the needle in real time without radiation or contrast and is portable and widely available.18 Fluoroscopic confirmation has shown that injectate delivered into the suprapatellar recess under ultrasound guidance reaches the joint, with a reported sensitivity of 96.3%.19

Direct comparisons of the two modalities are limited. A meta-analysis of glenohumeral injections found a higher, but not significantly different, accuracy with ultrasound than with fluoroscopy (93% vs 80%),20 a multicentre shoulder study reported comparable success with both techniques,21 and a non-randomized study of mixed joints found similar pain improvement after ultrasound-guided and fluoroscopy-guided injection.22 Randomized evidence for the knee, particularly from resource-limited settings, is scarce. We therefore compared the effectiveness of ultrasound-guided and fluoroscopy-guided intra-articular knee injection in relieving pain, measured on the visual analogue scale (VAS), over 2 weeks.

was obtained from every participant. Participants Patients referred from the Orthopaedics and Rheumatology departments for image-guided knee injection were screened consecutively. Men and women aged 18–70 years who required intra-articular knee injection for osteoarthritis, rheumatoid arthritis or another inflammatory knee condition were eligible. Patients were excluded if they were pregnant or had another contraindication to radiographic imaging; had joint infection or sepsis; had an anatomical deformity or prosthesis that could interfere with needle placement or imaging; had undergone total knee arthroplasty or another operation that altered knee anatomy; or had haemophilia or another coagulopathy. Age, sex, clinical indication, side, and duration of symptoms were recorded on a structured proforma. Randomization Participants were randomly allocated in equal numbers to Group A (ultrasound-guided injection) or Group B (fluoroscopy-guided injection. Each patient received a single injection by one technique only. Interventions All procedures were performed under strict aseptic technique by a single experienced operator, and all patients received the same injectate: 2 mL lidocaine (Xylocaine), 2 mL dexamethasone and 3–4 mL normal saline. Group A (ultrasound guidance). The patient lay supine with the knee flexed 20–30° over a pillow. A 7–12 MHz linear transducer was placed transversely over the superolateral aspect of the knee to show the suprapatellar recess deep to the quadriceps tendon. After skin preparation and local anaesthesia, a 22-gauge spinal needle was advanced in plane from lateral to medial under continuous visualization. Placement was accepted when the needle tip lay within the recess and the injectate was seen distending it. Group B (fluoroscopy guidance). The patient lay supine with the knee extended. Under intermittent fluoroscopy, a 22-gauge spinal needle was advanced through a superolateral approach toward the patellofemoral joint space. Intra-articular placement was confirmed by injecting 1–2 mL of iohexol (Omnipaque) and observing it outline the joint cavity, after which the drug was injected. Fluoroscopy time was kept to a minimum. Outcomes The primary outcome was pain intensity on a 10-cm VAS labelled in the patient’s native language and anchored at 0 (no pain) and 10 (worst imaginable pain). The same explanation script was used for all participants, and patients marked the scale independently at baseline (before the procedure), within 30 minutes after the procedure, and on days 3, 7 and 14. Pain was categorized as none (0), mild (1–3), moderate (4–6) or severe (7–10). Secondary outcomes were the change in VAS from baseline, the distribution of pain categories at day 14, and the proportion of patients achieving at least 50% reduction from baseline VAS at day 14. Sample size With 78 participants per group, the trial had 90% power at a two-sided significance level of 0.05 to detect a between-group difference of 1.0 point in mean VAS, assuming a standard deviation of 1.9 points. Statistical analysis Data were analysed with SPSS version 29.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation (SD) and categorical variables as number (percentage). Normality was assessed with the Shapiro–Wilk test. Between-group differences in mean VAS and in change from baseline were compared with the independent-samples t test and are reported with mean differences and 95% confidence intervals (CIs). Within-group change from baseline was assessed with the paired t test. Categorical variables were compared with the chi-square test, with Yates’ continuity correction for 2 × 2 tables. Age group, sex and indication were examined as potential effect modifiers by stratification, with the independent-samples t test applied within strata. All randomized patients were analysed in the group to which they were allocated. A two-sided p value < 0.05 was considered statistically significant.

A total of 156 patients were randomized, 78 to each group; all received the allocated injection, completed the 14-day follow-up and were included in the analysis (Figure 1). The groups were comparable at baseline in age, sex, indication, side, duration of symptoms and pain severity (Table 1). Osteoarthritis was the most common indication (78.8% overall), and most patients had severe pain at baseline (76.3%).

Figure 1. Flow of participants through the trial (CONSORT diagram). All 156 randomized patients received the allocated intervention, completed follow-up and were analysed.

Table 1. Baseline characteristics of the participants (N = 156)

SD, standard deviation; VAS, visual analogue scale (0–10). ^a Independent-samples t test. ^b Chi-square test. ^c Chi-square test with Yates’ continuity correction. No patient had no or mild pain at baseline.

Pain decreased significantly from baseline at every follow-up visit in both groups (paired t test, p < 0.001; Figure 2). Mean VAS did not differ significantly between groups at any time point (Table 2). At day 14, mean VAS was 3.81 ± 1.87 in the ultrasound group and 3.96 ± 1.72 in the fluoroscopy group (mean difference −0.15; 95% CI −0.72 to 0.41; p = 0.593). The mean reduction from baseline at day 14 was 3.62 ± 1.45 and 3.28 ± 1.31, respectively (mean difference 0.33; 95% CI −0.10 to 0.77; p = 0.134).

Figure 2. Mean visual analogue scale (VAS) pain scores over 14 days after ultrasound-guided (solid line, filled circles) and fluoroscopy-guided (dashed line, open squares) intra-articular knee injection. Error bars show 95% confidence intervals. Pain decreased significantly from baseline in both groups (p < 0.001), with no significant between-group difference at any time point.

Table 2. Visual analogue scale pain scores and reduction from baseline by group

CI, confidence interval; SD, standard deviation; VAS, visual analogue scale (0–10). ^a Ultrasound minus fluoroscopy; negative values for VAS scores and positive values for reductions favour ultrasound. ^b Independent-samples t test. Reduction from baseline was significant at every time point in both groups (paired t test, p < 0.001).

At day 14, the distribution of pain categories was similar in the two groups (p = 0.692; Table 3). No or mild pain was reported by 36 patients (46.2%) after ultrasound-guided injection and 31 (39.7%) after fluoroscopy-guided injection (p = 0.518), and a reduction of at least 50% from baseline VAS was achieved by 43 (55.1%) and 41 (52.6%) patients, respectively (p = 0.872). After stratification by age group, sex and indication, no stratum showed a significant difference between the techniques (all p > 0.05).

Table 3. Pain category and response at day 14 by group

VAS, visual analogue scale (0–10). ^a Chi-square test across the four categories. ^b Chi-square test with Yates’ continuity correction.

In this randomized trial, ultrasound-guided and fluoroscopy-guided intra-articular knee injections produced similar pain relief. Both techniques reduced mean VAS by more than 3 points, from about 7.3 at baseline to about 3.9 at 2 weeks, and no significant between-group difference was found at any time point or in any stratum. The size of the observed difference is more informative than its p value. At day 14 the mean difference was 0.15 points, and its 95% CI (−0.72 to 0.41) excluded a difference of 1 point in either direction. For comparison, the minimal clinically important improvement in pain for knee osteoarthritis has been estimated at about 20 mm on a 100-mm VAS.24 A clinically important advantage of either technique over 2 weeks is therefore unlikely, although the trial was not designed or powered as a formal equivalence study. A likely explanation is that both techniques placed the injectate within the joint in nearly all patients. Image-guided knee injections achieve accuracy above 95% in most series,9–12 fluoroscopy allows contrast confirmation of placement,17 and ultrasound-guided suprapatellar injection has been shown fluoroscopically to reach the joint.19 Once delivery is reliable, the drug rather than the guidance method determines the clinical effect. The main gain from imaging appears to come from moving from blind to image-guided injection; the trials that showed better outcomes with ultrasound compared it with palpation-guided injection, not with fluoroscopy.13–15 Our findings agree with the limited comparative evidence. In a prospective study of 71 patients with refractory synovitis, Furtado et al. found significant and similar improvement in pain after ultrasound-guided and fluoroscopy-guided injection,22 and a meta-analysis of glenohumeral injections found no significant difference in accuracy between the two modalities.20 Ertilav et al. also reported significant improvement in VAS after ultrasound-guided knee injection using either suprapatellar or infrapatellar approaches, with placement confirmed fluoroscopically.25 Because effectiveness was comparable, other considerations should guide the choice of technique. Ultrasound involves no ionizing radiation for the patient or operator and no iodinated contrast; it is portable, can be used in an outpatient room, and shows effusion, synovitis and soft tissues that fluoroscopy cannot.16,18,21 In our setting, ultrasound-guided injections were performed in the outpatient ultrasound room, whereas fluoroscopy required the fluoroscopy suite. Given the growing burden of knee osteoarthritis in Pakistan,2,3 these practical advantages favour ultrasound for routine injections. Fluoroscopy remains a valid option when ultrasound expertise is unavailable or sonographic visualization is difficult. Limitations This study has limitations. Follow-up was limited to 2 weeks, so the durability of relief was not assessed. The trial was conducted at a single centre, which may limit generalizability. Patients and the operator could not be blinded because the procedures differ visibly, and pain was self-reported. All injections were performed by one experienced operator, so results may differ with less experienced sonologists. Intra-articular placement in the ultrasound group was judged sonographically without contrast confirmation. The injectate contained a local anaesthetic, which influences pain within the first 30 minutes. Procedure time, radiation dose and costs were not measured, so the practical advantages of ultrasound could not be quantified. Finally, the sample was not powered to detect differences smaller than 1 VAS point.

Ultrasound-guided and fluoroscopy-guided intra-articular knee injections were equally effective in reducing pain over 2 weeks, with no significant difference at any time point. Given its freedom from ionizing radiation and contrast, lower resource requirements and bedside availability, ultrasound can be used as an effective alternative to fluoroscopy for intra-articular knee injection.

DECLARATIONS

Ethics approval and consent to participate: Approved by the Institutional Review Board of King Edward Medical University, Lahore. Written informed consent was obtained from all participants.