Table 4.
Extended lumbar drainage
| Study | Shunted patients | Methodology | Criteria for + ve ELD | Criteria for SR | Side effects | Main reported outcomes |
|---|---|---|---|---|---|---|
| Gallina et al. (2018) [19] | n = 68 |
• Assessments: MMSE, objective urinary incontinence and gait scale (4 categories from functional to completely dysfunctional) • 24-h ELD • Follow-up: 12 months |
• + 2 points (urinary incontinence + gait scale) or + 1 on urinary incontinence scale or gait scale and minimum 3 points on MMSE) | • + 2 points (urinary incontinence + gait scale) or + 1 on urinary incontinence scale or gait scale and minimum 3 points on MMSE) | • Intracranial hypotension (n = 1), root irritation (n = 2), headache (9.9%), overall procedural complications (2.1%) |
• 73.3% of patients had a positive outcome to 1-day ELD • ELD had a sensitivity = 100%, specificity = 75.0%, PPV = 96.8%, NPV = 100% • TP = 60, FN = 0 FP = 2, TN = 6 |
| Marmarou et al. (2005)[38] | n = 102 |
• Assessments: MMSE, gait and bladder function, and neuro-psychometric parameters • 3-day ELD • A shunt was offered based on ELD response or patient request • Follow-up: 12 months |
• A 10-day survey by patients and caregivers stating improvement in clinical status | • The same criteria used to assess 3-day ELD response were used to assess surgical outcome | • Infection in 2 (1.3%) of 151 patients, and 4 patients (2.6%) experienced headache |
• There was a statistically significant correlation between ELD responders and shunt responsiveness (p < 0.0001) • ELD prediction of shunt responsiveness has a sensitivity of 95% (CI: 84–90%), specificity of 64% (CI: 44–84%), PPV of 90% (CI: 72–100%), NPV of 78% (CI: 70–80%) • TP = 76, FN = 4, FP = 8, TN = 14 |
| Chotai et al. (2014) [8] | n = 60 |
• Assessments: videography of gait assessment, balance, muscle strength, speech fluency, behavior, and MMSE • A 4-day ELD • Follow-up: 12 months |
• Increase of ≥ 2 points on MMSE or improvement in gait, balance speech fluency was assessed based on videotape assessment and documentation |
• Gait and cognitive improvement (MMSE improvement of 2 >) were the primary outcomes • Functional outcomes were assessed using a survey by the family and patient |
• 10% of ELD patients experienced. Transient nerve root irritation |
• A statistically significant improvement in cognition on day 4 following ELD was observed with median MMSE score increasing to 27 from 23.5 (χ2 = 15.74, p = 0.001), with no improvement in gait • 4-day ELD: sensitivity = 100%, specificity = 60%, NPV = 100%, and PPV = 96% • ROC analysis demonstrated reasonable accuracy for ELD prediction of SR (area under curve = 0.8 ± 0.14, p = 0.02, CI = 0.52–1.0) • TP = 55, FN = 0, FP = 2, TN = 3 |
| Mahr et al. (2016) [37] | n = 31 |
• Assessments: Kiefer score, SLHS, MMSE, mRS standardized gait testing, and grooved pegboard test • A 3-day ELD • Follow-up: 12 months |
• A 10% improvement in gait or a 10% improvement in MMSE • Qualitative assessment by patient and family also considered |
• A ∆Kiefer values 12-month post-shunt was used to assess SR with patients categorized as excellent responders (relief of all symptoms), improved patients (∆Kiefer reduction of at least 10%), and non-responders (∆Kiefer reduction < 10%) | • Not reported | • ELD response predicted improvement post-shunt surgery in 87.9% of patients with iNPH |
| Eide and Stanisic (2010) [17] | n = 31 |
• Assessments: gait analysis and NPH scale • 3-day ELD was used with ICPM • Follow-up: 3 and 6–12 months |
• Gait function was assessed using a video analysis with an improvement in gait function as a positive ELD response | • Increase + ≥ 2 scores on their NPH scale | • Not reported |
• ELD responders and non-responders had significantly different ICP wave amplitudes (p < 0.001) • All patients (53.6%) with elevated pulsative ICP had a clinical response to ELD, compared to 23.1% of the low ICP group (PPV = 100% and NPV = 77% for clinical response to ELD) • The reduction in ICP wave amplitude during ELD was related to the changes in NPH scores (Spearman correlation − 0.6; p < 0.001) after shunt treatment • TP = 15, FN = 1, FP = 3, TN = 2 |
| Panagiotopoulos et al. (2005) [47] | n = 22 |
• Assessments: history taking, neurological exam, MMT and NPH score • 5-day ELD • Follow-up: 3 month |
• + 1 or more points on the NPH scale | • Increase in NPH score | • None reported |
• In patients able to walk, improvement after ELD in gait disturbance was significantly correlated with improvement 3-month post-shunting (Pearson’s r = 0.833, p < 0.01) • Quantitative NPH score analysis for 3-month post-shunt correlated to an improvement after ELD (Spearman’s rho = 0.462, p = 0.03) • TP = 9, FN = 2, FP = 0, TN = 11 |
| Chaudhry et al. (2007) [7] | n = 60 |
• Assessments: RAVLT, Boston naming test, COWA, Wechsler memory logical memory test, alphabet writing, line tracing and coping pentagons, Rey complex figure test and grooved pegboard test • A 2–3-day ELD • Follow-up: 3 months |
• 1 SD or more improvement in RAVLT/WMS | • 1 SD or more improvement in RAVLT/WMS | • None reported |
• In 3 subsets evaluating learning, retention and delayed recall of the RAVLT, the magnitude of improvement post-ELD insertion was predictive of the magnitude of improvement after shunt surgery: learning (r 2 = 0.58; p < 0.001), retention (r2 = 0.32; p = 0.04), and delayed recall (r 2 = 0.36; p = 0.02) • A 5- or more point improvement on RAVLT post-drainage was associated with a significant improvement on > 0.5 the memory tests post-VPS (chi-squared = 10.8; p = 0.0005) and had PPV = 50% and NPV = 96% |
| Woodworth et al. (2009) [71] | n = 51 |
• Assessments: objective symptom analysis • Controlled continuous CSF drainage at 10 mL/h (240 mL/day) for 3 days • Follow-up: 1, 3, 6, 12 months and yearly thereafter |
• ≥ 1 iNPH symptom improvement • OR probable B-waves present during Pcsf monitoring |
• Objective and family reported improvement in 1, 2, or 3 iNPH symptoms | • None reported |
• Continuous lumbar drainage prediction of SR: sensitivity = 91%, specificity = 70% • Patients with a positive response to CSF drainage were 3.2 times more likely to improve following CSF shunting (RR = 3.2; [CI: 0.09–1.00], p < 0.05) • A positive CSF drainage response predicted VPS responsiveness (RR = 0.30, [CI: 0.09–0.98], p < 0.05) |
Studies included assessing the use of extended lumbar drainage (ELD) in predicting shunt responsiveness.
COWA, controlled oral word association test; CSF, cerebrospinal fluid; ICP, intracranial pressure; iNPH, idiopathic NPH; MMT; mini mental test; MMSE, mini mental state exam; mRS, modified Rankin scale; NPH, normal pressure hydrocephalus; Pcsf, CSF pressure; TT, tap test; VPS, ventriculoperitoneal shunt; RAVLT, Rey auditory visual learning test; SD, standard deviation; SLHS, Stein and Langfitt hydrocephalus score; SR, shunt response; WMS, Wechsler memory scale.