. 2009 Sep 15;19(2):181–204. doi: 10.1007/s00586-009-1155-x

Table 7.

Retrospective cohort studies

Study	Main inclusion criteria, main exclusion criteria	Type /level LDH	Interventions/technique/instrumentation	Follow-up: duration and outcome	Comment
Ahn et al. [3]	Inclusion criteria	n = 43 ♀11 ♂32, mean 46 years, range 22–72	Percutaneous endoscopic lumbar discectomy (PELD)	Follow-up: range 24–39 months, 0% non-responders	Authors included only patients with recurrent LDH, more than 6 months after open microdiscectomy
	Prior disc surgery	Type: all LDH	Intradiscal and intracanal technique, instrumentation not specified	Pain (VAS): pre-op. 8.7, follow-up 2.6, difference 6.1 = 70%
	Radiculopathy	Level: single level, L3–S1		GPE (MacNab): 28% excellent, 53% good, 4.7% poor
	Post-tension sign	Level: single level, L3–S1		Complications: 4.6%
	Neurological deficit			Re-operations: 2.3%
	Exclusion criteria
	Segmental instability
	Spondylolisthesis
	Calcified fragments
Chiu [5]	Inclusion criteria	n = 2,000 ♀990 ♂1010, mean 44 years, range 24–92	Transforaminal microdecompressive endoscopic assisted discectomy (TF-MEAD)	Follow-up: mean 42 months (range 6–72), 0% non-responders	Authors included also patients with stenosis and degenerative disc disease
	Virgin and prior disc surgery	Type: not specified	Intradiscal and intracanal technique	GPE (unclear instrument): 94% excellent or good, 3% poor
	Pain in back	Level: single and multiple level	Karl Storz instrumentation	Complications: 1%
	Radiculopathy			Re-operations: not specified
	Neurological deficit
	Exclusion criteria
	Cauda syndrome
	Painless motor deficit
Choi et al. [6]	Inclusion criteria	n = 41, ♀23 ♂18, mean 59 years, range 32–74	Extraforaminal targeted fragmentectomy	Follow-up: mean 34 months (range 20–58), 4.9% non-responders
	Radiculopathy	Type: extraforaminal LDH	Pure intradiscal technique, YESS, Richard Wolf instrumentation	Pain leg (VAS): pre-op. 8.6, follow-up 1.9, difference 6.7 = 78%
	Post-tension sign	Level: single level, L4–S1		Return to work: mean 6 weeks (range 4–24)
	Neurological deficit	Level: single level, L4–S1		Functional status (ODI): pre-op. 66.3, follow-up 11.5, difference 54.8 = 83%
	Exclusion criteria			PS: 92%
	Previous surgery (same level)			Recurrence: 5.1%
	Central or lateral stenosis			Complications: 5.1%
	Segmental instability			Re-operations: 7.7%
	Calcified disc
Ditsworth [7]	Inclusion criteria	n = 110 ♀40 ♂70, median 55 years, range 20 to > 60	Endoscopic transforaminal lumbar discectomy	Follow-up: range 24–48 months, 0% non-responders
	Radiculopathy	Type: all LDH	Intradiscal and intracanal technique	GPE (MacNab): 91% excellent or good, 4.5% poor
	Post-tension sign	Level: single level	Flexible endoscope	Recurrence: 0%
	Neurological deficit			Complications: 0.9%
	Exclusion criteria			Re-operations: 4.5%
	Spinal stenosis
	Segmental instability
Eustacchio [8]	Inclusion criteria	n = 122 ♀36 ♂86, median 55 years, range 18–89	Endoscopic percutaneous transforaminal treatment	Follow-up: mean 35 months (range 15–53), 0% non-responders	Authors excluded n = 10 from analyses due to stopped procedures. These were taken into account in this review
	Radiculopathy	Type: all LDH	Intradiscal and intracanal technique instrumentation not specified	GPE (MacNab): 45% excellent, 27% good, 27% poor
	Post-tension sign	Level: multiple level n = 4, L2–S1		Functional status (PROLO): 71.9% excellent or good
	Neurological deficit	Level: multiple level n = 4, L2–S1		Return to work: 94%
	Exclusion criteria			Recurrence: 12%
	Cauda syndrome			Complications: 9%
				Re-operations: 27%
Haag [10]	Inclusion criteria	n = 101	Transforaminal endoscopic microdiscectomy	Follow-up: mean 28 months (range 15–26), 9% non-responders	Authors excluded n = 3 from analyses due to technical problems during procedures. These were taken into account in this review
	Radiculopathy	Type: all LDH	Pure intradiscal technique	PS: good: 66%, satisfied: 9%, poor: 25%
	Neurological deficit	Level: single level, L2–S1	Sofamor–Danek instrumentation	Complications: 7.6%
	Exclusion criteria	Level: single level, L2–S1		Re-operations: 17%
	Discus narrowing
	Calcified disc
Hochschuler [13]	Inclusion criteria	n = 18 ♀5 ♂13, mean 31 years, range 18–55	Arthroscopic microdiscectomy (AMD)	Follow-up: mean 9 months (range 4–13), 0% non-responders
	Radiculopathy	Type: not specified	Pure intradiscal technique	Re-operations: 11%
	Exclusion criteria	Level: L3−S1	Kambin technique
	Previous operation (same level)
	Sequestration
	High iliac crest
Hoogland [14]	Inclusion criteria	n = 246	Transforaminal endoscopic discectomy with foraminoplasty	Follow-up: 24 months, 0% non-responders	Authors included also patients with foraminal stenosis
	Not specified	Type: not specified	Intracanal technique, Thessys instrumentation	GPE (MacNab): 86% excellent or good, 7.7% poor
	Exclusion criteria	Level: not specified		Complications: 1.2%
	Not specified			Re-operations (1st year): 3.5%
Iprenburg [18]	Inclusion criteria	n = 149 ♀62 ♂87, mean 43 years, range 17–82	Transforaminal endoscopic surgery	Follow-up: not specified, 29% non-responders
	Not specified	Type: all LDH	Intracanal technique, Thessys instrumentation	Pain (VAS): not specified
	Exclusion criteria	Level: single level, L3–S1		Functional status (ODI): not specified
	Central stenosis	Level: single level, L3–S1		Recurrence: 6%
				Complications: not specified
				Re-operations: not specified
Jang et al. [19]	Inclusion criteria	n = 35 ♀20 ♂15, mean 61 years, range 22–84	Transforaminal percutaneous endoscopic discectomy (TPED)	Follow-up: mean 18 months (range 10–35), 0% non-responders
	Radiculopathy	Type: foraminal and extraforaminal LDH	Intradiscal and intracanal technique, instrumentation not specified	Pain (VAS): pre-op. 8.6, follow-up 3.2, difference 5.4 = 63%
	Exclusion criteria	Level: single level, L2–S1		GPE (MacNab): 86% excellent or good, 8.6% poor
	Previous surgery (same level) segmental instability	Level: single level, L2–S1		Recurrence: 0%
	Spinal stenosis			Complications: 17%
	Listhesis			Re-operations: 8.6%
Lew et al. [35]	Inclusion criteria	n = 47 ♀12 ♂35, mean 51 years, range 30–70	Transforaminal percutaneous endoscopic discectomy	Follow-up: mean 18 months (range 4–51), 0% non-responders
	Radiculopathy	Type: foraminal and extraforaminal LDH	Pure intradiscal technique	GPE (MacNab): 85% excellent or good, 11% poor
	Post-tension sign	Level: L1–L5	Surgical dynamics instrumentation	Return to work: 89%
	Neurological deficit			Complications: 0%
	Exclusion criteria			Re-operations: 11%
	Previous surgery (same level)
Mayer and Brock [39]	Inclusion criteria	n = 30 ♀11 ♂19	Percutaneous endoscopic lumbar discectomy (PELD)	Follow-up: range 6–18 months, 0% non-responders	Twenty of the patients were described in a prospective study [41]. In this review reoperations were labelled as moderate or poor outcome on GPE
	Radiculopathy	Type: not specified	Pure intradiscal technique, instrumentation not specified	GPE (S/S-score): 67% excellent or good, 33% moderate or poor
	Post-tension sign	Level: multiple level n = 1, L2–L5		Return to work: 7.1 ± 4.2 weeks, 90% (6 months)
	Neurological deficit	Level: multiple level n = 1, L2–L5		Complications: 3.3%
	Exclusion criteria			Re-operations: 3.3%
	Sequestration
	Previous surgery (same level)
	Cauda syndrome
	Segmental instability
	Spinal stenosis
	Listhesis
Savitz [49, 50]	Inclusion criteria	n = 300 ♀132 ♂168, range 16–81 years	Percutaneous lumbar discectomy with endoscope	Follow-up: 6 months, 0% non-responders
	Radiculopathy	n = 300 ♀132 ♂168, range 16–81 years	Percutaneous lumbar discectomy with endoscope	Follow-up: 6 months, 0% non-responders
	Post tension sign	Type: not specified	Pure intradiscal technique, Kambin technique	Return to work (6 months): 67%
	Neurological deficit	Type: not specified	Pure intradiscal technique, Kambin technique	Return to work (6 months): 67%
	Exclusion criteria	Level: multiple level n = 40, L2–S1		Complications: 5.3%
	Previous surgery (same level)			Complications: 5.3%
	Sequestration			Re-operations: 1.3%
	Obesity			Re-operations: 1.3%
Schreiber and Suezawa [53]; Suezawa and Schreiber [58]; Leu and Schreiber [36]; Schreiber and Leu [52]	Inclusion criteria	n = 174 ♀68 ♂106, mean 39 years, range 16–81	Percutaneous nucleotomy with discoscopy	Follow-up: mean 28 months, 0% non-responders	Authors included also patients with degenerative disc disease, only the scores from LDH are quoted in this review
	Radiculopathy	Type: not specified	Pure intradiscal technique	GPE (S/S-score): 85% excellent or good
	Exclusion criteria	Level: multiple level n = 25	Modified Hijikata instrumentation biportal	Complications: 10%
	Sequestration			Re-operations: 21%
Shim et al. [56]	Inclusion criteria	n = 71 ♀39 ♂32, mean 45 years, range 21–74	Transforaminal endoscopic surgery	Follow-up: mean 6 months (range 3–9), 0% non-responders	n = 14 patients with L5−S1 level LDH are operated via a interlaminar approach
	Radiculopathy	Type: not specified	Pure intradiscal technique	GPE (MacNab): 33% excellent, 45% good, 6.5% poor
	Exclusion criteria	Level: single level, T12–S1	YESS, Richard Wolf instrumentation	Complications: 2.8%
	Not specified	Level: single level, T12–S1		Re-operations: 7.0%
Tsou and Yeung [59]	Inclusion criteria	n = 219 ♀83 ♂136, mean 42 years range 17–71	Transforaminal endoscopic decompression	Follow-up: mean 20 months (range 12–108), 11.9% non-responders	Possible patient overlap with other study [65]
	Radiculopathy	Type: central LDH	Intradiscal and intracanal technique	GPE (MacNab): 91% excellent or good, 5.2% poor
	Neurological deficit	Level: single level, L3–S1	YESS, Richard Wolf instrumentation	Recurrence: 2.7%
	Exclusion criteria	Level: single level, L3–S1		Complications: 2.7%
	Sequestration			Re-operations: 4.6%
	Previous operation (same level)
Tzaan [60]	Inclusion criteria	n = 134 ♀56 ♂78, mean 38 years, range 22–71	Transforaminal percutaneous endoscopic lumbar discectomy (TPELD)	Follow-up: mean 38 months (range 3–36), 0% non-responders
	Pain in leg and back	Type: all LDH	Pure intradiscal technique	GPE (modified MacNab): 28% excellent, 61% good, 3.7% poor
	Exclusion criteria	Level: multiple level n = 20, L2–S1	Instrumentation not specified	Recurrence: 0.7%
	Sequestration	Level: multiple level n = 20, L2–S1		Complications: 6.0%
	Spinal stenosis			Re-operations: 4.5%
	Calcified disc
	Segmental instability
	Cauda syndrome
Wojcik [62]	Inclusion criteria	n = 43 ♀25 ♂18, mean 30 years	Endoscopically assisted percutaneous lumbar discectomy	Follow-up: 18 months, 16.3% non-responders
	Radiculopathy	Type: not specified	Pure intradiscal technique	GPE (unclear instrument): 64% good, 36% satisfied, 0% poor
	Exclusion criteria	Level: not specified	Modified Hijikata instrumentation	Complications: not specified
	Sequestration			Re-operations: not specified
	Chronic back pain
Yeung and Tsou [65]	Inclusion criteria	n = 307 ♀102 ♂205, mean 42 years, range 18–72	Posterolateral endoscopic excision for lumbar disc herniation	Follow-up: mean 19 months (range 12–?), 8.8% non-responders	Possible patient overlap with other study [65]
	Prior disc surgery n = 31	Type: all LDH	Intradiscal and intracanal technique	GPE (MacNab): 84% excellent or good, 9.3% poor
	Radiculopathy	Level: single level, L2–S1	YESS, Richard Wolf instrumentation	Recurrence: 0.7%
	Neurological deficit	Level: single level, L2–S1		Complications: 3.9%
	Exclusion criteria			Re-operations: 4.6%
	Sequestration
	Central and lateral stenosis

? unknown, is not described in the study

Intervention as quoted in original article, Post-tension signs denotes positive tension signs (straight leg raising test or contralateral straight leg raising test)

Outcomes: S statistically significant, NS not statistically significant, PS patient satisfaction, MacNab MacNab score as described by MacNab [39]. The sum of ‘excellent’ and ‘good’ outcomes are labelled ‘satisfactory’, GPE global perceived effect, S/S-score Suezawa and Schreiber score [40], ODI Oswestry disability index [38], PROLO prolo functional-economic outcome rating scale [44]