Current practice patterns among Indian neurologists in the Evaluation and management of Normal Pressure Hydrocephalus: A nationwide cross‐sectional survey

Abstract

INTRODUCTION

This cross‐sectional study explores the diagnostic and management practices of normal pressure hydrocephalus (NPH) in India.

METHODS

A 16‐question online survey based on the Checklist for Reporting Results of Internet E‐Surveys checklist was disseminated via closed social media and academic groups.

RESULTS

A total of 59 neurologists responded. There was significant variability in how the cerebrospinal fluid (CSF) tap test (TT) was implemented. Although 95% of practitioners performed a CSF TT before referring patients for surgery, there were differences in the volume of CSF removed, the assessment tools used, thresholds, and the timing of post‐test evaluations. Notably, 36% of neurologists relied on subjective assessment instead of objective scoring to determine CSF TT responsiveness. One‐third still considered surgical referral even if the CSF TT was negative, especially when imaging was strongly suggestive of the diagnosis.

CONCLUSION

This study highlights considerable heterogeneity in the evaluation of NPH in India. We need evidence‐based practice parameters to ensure accurate diagnosis, informed treatment decisions, and improved patient outcomes.

Highlights

• There is a significant variability in clinical practice globally, particularly regarding the use of the cerebrospinal fluid tap test (CSF TT).

• Before referring for surgery, 95% of practitioners did a CSF TT.

• Nearly one‐third of neurologists relied on subjective clinical assessment instead of objective scores.

• One‐third considered surgical referral, even if the CSF TT was negative, especially when imaging was strongly suggestive.

• There is a need for consensus‐based practice parameters to ensure accurate diagnosis, informed treatment decisions, and improved patient outcomes.

1. BACKGROUND

Normal pressure hydrocephalus (NPH) is characterized by a specific triad of symptoms, including gait disturbance, cognitive disorders, and urinary incontinence. Idiopathic NPH (iNPH) affects 0.5% to 2.9% of the elderly population. Japanese guidelines classify iNPH as “possible iNPH,” “probable iNPH,” or “definite iNPH.” ¹ According to these guidelines, the diagnosis of definite iNPH is made when objective improvement of symptoms is demonstrated after the shunt surgery. However, deciding to subject a patient to surgery is based on the clinico‐radiologic profile and ancillary cerebrospinal fluid (CSF) dynamic tests, such as a CSF tap test (CSF TT), external lumbar drainage, or lumbar infusion test.

The practice varies between regions. According to Japanese Guidelines, a positive CSF TT is not necessary to refer patients for surgery. In contrast, in various parts of the world, a definite improvement after the CSF TT often gives both the physician and the patient greater confidence. In the algorithm used at Imperial College, London, UK, a negative CSF TT is followed by a 48‐h lumbar drainage or a lumbar infusion test, and surgical treatment is offered if either test is positive. ² In Canadian health care, only patients with a positive CSF TT are offered surgery. ³ However, there is wide variation in clinical practice. In the United States, analysis of the National Inpatient Database revealed regional differences in the utilization of lumbar puncture, external ventricular drains, and CSF‐diversion methods, such as lumboperitoneal or ventriculoperitoneal shunting. ⁴

Because gait, cognition, and urinary symptoms are relatively common in older adults, we need evidence‐based practice parameters to evaluate those with ventriculomegaly on imaging. Although widely used, the CSF TT exhibits significant variation in the volume of CSF tapped, the parameters assessed after tapping, and the timing of their assessment, as shown in a recent systematic review. ⁵ However, there is no literature on the variability among neurologists in their evaluation strategies. Based on anecdotal observations and discussions with peers by the first author, although the guidelines do not suggest a positive TT, practitioners in India seem to refer only CSF TT responders for surgery, probably because they feel it increases the probability of good postoperative outcomes. In addition, several practitioners feel that they may get into medico‐legal issues if CSF TT non‐responders are referred for surgical intervention. Our study aims to explore current practice trends among Indian neurologists in defining CSF TT responsiveness and to understand the diagnostic tests used in the pre‐surgical evaluation of NPH, the protocol used for CSF TT, cutoffs, and tools used to determine CSF TT responsiveness, as well as surgical referral patterns, using an online questionnaire. We aimed to explore the practice settings of the survey respondents, the number of patients they see, and the clinical tests and cutoffs used to select patients with iNPH for surgical treatment.

2. METHODS

2.1. Survey design

RESEARCH IN CONTEXT

1. Systematic review: The authors reviewed the existing literature regarding the evaluation of idiopathic normal pressure hydrocephalus (iNPH) using the cerebrospinal fluid (CSF) tap test (TT). In the systematic review by Bluett et al., significant inconsistencies were found in the protocol for performing and interpreting the CSF TT. We reviewed the existing guidelines, and the Japanese guidelines suggest operating even if the CSF TT was negative, if other clinical and imaging features are suggestive of NPH. On the other hand, Canadian and UK practices relied on the CSF TT to guide patient selection for ventriculoperitoneal (VP) shunt.

2. Interpretation: This nationwide survey provides the first systematic assessment of iNPH practice patterns among Indian neurologists. We identify substantial variability in diagnostic pathways, use of CSF TT, imaging interpretation, and referral for shunt surgery. The findings reveal a gap between published recommendations and real‐world decision‐making, shaped by uncertainty and health‐system constraints.

3. Future directions: Variability in diagnostic practices and referral thresholds for surgery may create significant barriers to the systematic study of the disease. India‐specific consensus pathways, training, and implementation strategies are urgently needed. This study establishes a policy‐relevant baseline to guide guideline adaptation, training, and system‐level reforms in resource‐diverse settings.

Based on a thorough literature review, we designed the survey to address the existing knowledge gaps. The study was designed in line with the Checklist for Reporting Results of Internet E‐Surveys checklist for online surveys. ⁶ The survey questions were developed by the first (S.P.) and the corresponding author (A.E.). These questions were included based on a literature review and to address identified knowledge gaps. Face and content validity were assessed by a panel of movement disorder neurologists who were not survey respondents. The questionnaire consisted of 16 questions (Table 1, Q1–Q16). All questions were mandatory for the survey to be completed and submitted, ensuring a balanced picture of the evaluation practices. The completeness check feature of Google Forms (Google LLC, Mountain View, CA, USA) facilitated this. Three questions pertained to the practice setting and the number of patients evaluated and referred for surgery over a 1‐year period. We had one question regarding the threshold for initiating diagnostic evaluation, 10 questions regarding the CSF TT and the scales used for assessment, ⁷ and two questions regarding magnetic resonance imaging (MRI) CSF flow studies. We were aware that none of the major centers performed external lumbar drainage or lumbar infusion testing, so we did not have questions about them. We wanted to keep the questionnaire short to ensure a better response rate. The questions were in multiple‐choice format. This was followed by informal pre‐testing by the study team, that is, the survey was thoroughly tested for ease of use and technical functionality. The average time to complete the survey was 3 to 5 min, as assessed during the pre‐testing. The study and the survey questionnaire were approved by the Institutional Review Board of the All India Institute of Medical Sciences, New Delhi, India.

TABLE 1.

Results of the survey: Practice trends of Indian neurologists in the evaluation and management of NPH.

Parameter	Response, n (%)
Number of respondents	59 (100)
Level of practice
Tertiary care level of practice	44 (75)
Independent clinics/secondary care	15 (25)
Number of patients of NPH seen per month
0–2 per month	48 (81.4)
3–5 per month	10 (16.9)
>5 per month	1 (1.7)
When do you evaluate for NPH? If the patient has
one of gait impairment, urinary symptoms, or cognitive impairment	17 (28.8)
Two of 3 symptoms (gait impairment, urinary symptoms, or cognitive impairment	22 (37.3)
“Triad” of gait impairment, urinary symptoms, or cognitive impairment	15 (25.4)
If imaging is suggestive of NPH, even if none of the symptoms of gait impairment, urinary symptoms, or cognitive impairment are present	5 (8.5)
Do you perform a CSF tap test before referring a patient for surgery?
I do a tap test on every patient before referring for surgery	56 (94.9)
I routinely do lumbar infusion test/external lumbar drainage for up to 72 h before referring for surgery	1 (1.7)
If CT/MRI structural imaging is characteristic, I may skip the CSF tap test and refer for shunt surgery	1 (1.7)
I will rely on CSF flow studies on MRI rather than the CSF tap test if other clinical features are consistent	1 (1.7)
How much CSF do you remove during the tap test?
<10 mL	0
10 mL	1 (1.7)
11–30 mL	15 (25.4)
31–50 mL	38 (64.4)
>50 mL	5 (8.5)
What are the various tests you use during the CSF tap test to diagnose NPH? (You can select multiple options.)
Boon's Gait Scale (composite of walk score, time score, and step score)	28(47.5)
iNPH grading scale (composite of Gait, cognition, urinary symptoms)	16 (27.1)
TUG test	22 (37.3)
Mini‐Mental State Examination (MMSE)	36 (61)
Montreal Cognitive Assessment (MoCA)	21 (35.6)
Subjective patient/clinician assessment of improvement	37 (62.7)
What is the cutoff you take in the Boon's Gait Scale (composite of walk score, time score, and step score) for CSF tap test responsiveness? (out of 40)
3	1(1.7)
3–5	17 (28.8)
I do not use cutoffs	14 (23.7)
I do not use the Boon's Gait Scale	27 (45.8)
What is the cutoff you take in MoCA/MMSE for CSF tap test responsiveness? (out of 30)
1‐point change	6(10.2)
2‐point change	10 (16.9)
3‐point change	19 (32.2)
4‐point change	7(11.9)
5‐point change	7(11.9)
I do not use MMSE/MoCA during the CSF tap test	10 (16.9)
What is the cutoff you take in the iNPH scale for CSF tap test responsiveness? (total 12 points)
1‐point change	3 (5.1)
2‐point change	13 (22)
3‐point change	8 (13.6)
I do not use the iNPH scale	35 (59.3)
What is the cutoff you take for the TUG test during the CSF tap test?
I use percentage cutoffs—more than a 10% reduction in time as a positive tap test	19 (32.2)
I use absolute cutoffs—more than a 5‐s reduction—as a positive tap test	6 (10.2)
I do not use the TUG test	34 (57.6)
What time after the tap test do you consider tap test responsiveness based on the cutoff you use? (You can select multiple options.)
1 h	28(4)
2 h	14(24)
4 h	11 (19)
6 h	34(58)
12 h	30 (51)
24 h	46 (78)
48 h	17 (29)
When will you consider the CSF tap test to be positive?
If at least at one time point, at least one score suggests improvement	34 (57.6)
If there is improvement in all scores and all symptoms	4 (6.8)
I pay attention only to patient/clinician subjective global improvement, not the percentage improvement in scores	21 (35.6)
What do you do if the CSF tap test does not lead to improvement in gait, cognition, or urinary symptoms or scores?
I will not refer for shunt surgery	39 (66.1)
I use it to prognosticate patients, but not to defer or refer for surgery	10 (16.9)
I still refer patients for surgery even if CSF tap test studies are negative, if imaging is highly suggestive of the diagnosis	10 (1.9)
Do you routinely use an MRI CSF flow study to diagnose NPH?
Yes	26(44)
No	33 (56)
What do you do if the MRI CSF flow study is not suggestive of NPH?
I will not refer for shunt surgery	20(34)
I use it only to prognosticate patients, not to defer or refer for surgery	27(46)
I still refer patients for surgery even if MRI CSF flow studies are negative	12(20)
How many patients do you refer for shunt surgery in a year? (Approximately)
0 to 2	21 (35.6)
3 to 5	27 (45.8)
6 to 10	8 (13.6)
More than 10	3 (5.1)

Note: The questionnaire was tested for readability, and Grammarly (Grammarly Inc., San Francisco, CA, USA) was used to check correctness, engagement, and clarity.

Abbreviations: CSF, cerebrospinal fluid; NPH, normal pressure hydrocephalus; TUG, timed up and go test.

2.2. Survey dissemination

We sent out an open survey using a self‐administered questionnaire to assess evaluation practices in the management of NPH, administered via Google Forms (Google LLC, Mountain View, CA, USA). The survey was circulated in closed professional groups on social media platforms, including WhatsApp groups of general and movement disorders–trained neurologists used for case‐based discussions and academic threads, as well as email groups of academic institutions, with a reach of ≈600 neurologists across various forums, including academic practitioners, community practitioners, and neurologists specializing in movement disorders. The responses were collected from November 2024 to March 2025. The survey forms mentioned on the opening page that the survey would be de‐identified for analysis and would be completely confidential and anonymous. The survey was not advertised, and the participants were not offered any incentive to respond.

The respondents were informed about the purpose of the study and the expected duration of the survey, and they provided consent before beginning the survey. The survey participation was entirely voluntary. We ensured that no duplicate responses were submitted by making it a closed survey. This was done by limiting the survey response to one per participant on the Google form. The survey was never displayed again after the user had filled it out. The login details were stored together with the survey results and later removed during the de‐identification process before data cross‐checking. Survey respondents could edit their responses during submission, but not after. The survey results were stored in secure, encrypted storage systems before analysis.

2.3. Statistical analysis

The survey responses were imported into a Microsoft Excel spreadsheet (Microsoft Corporation, Redmond, WA, USA) and cross‐verified for accuracy. We used Stata version 17.0 (StataCorp LLC, College Station, TX, USA) to summarize the results. We reported the descriptive statistics; inferential statistics did not apply to this data set. The summary statistics were presented numerically, with proportions expressed as percentages. When more than one option could be selected as a response, each option was considered separately, and proportions were calculated as either null or yes for each option.

3. RESULTS

We received 59 responses in total. Although the number of recipients is unknown, based on an estimated maximum reach of 600 neurologists, the response rate is ≈10%. The survey had a margin of error of ≈12%, assuming a reach of 600 neurologists involved in NPH care.

Most respondents practiced in tertiary care centers (75%), reflecting the level of care where patients with NPH are typically treated, as they are often treated with a ventriculoperitoneal shunt, which is a procedure typically performed at the tertiary care level. All the respondents were seeing and treating patients with NPH, although, in some months, they had seen no patients. The majority evaluated fewer than two patients with NPH per month (81.4%).

There was considerable variability in the threshold for initiating evaluation for NPH: 28.8% evaluated patients even if only one symptom from the triad was present, 37.3% required two symptoms, and 25.4% required the complete triad. Before referring the patients for surgery, 95% of respondents performed a CSF TT. The respondents differed in the volume of CSF removed, and in the thresholds used to define clinically meaningful improvement. Almost two‐thirds of participants drained 31–50 mL of CSF. The survey demonstrated substantial heterogeneity not only in the performance of the CSF TT but also in the methods used to assess post‐tap improvement. Although 62% of respondents used subjective assessments to assess response to the CSF TT, most also used some or the other scale to assess improvement. However, five respondents just used subjective measures and no scales to assess improvement following the TT. In addition, there was considerable variability in the timing of reassessment after CSF removal, with most practitioners evaluating patients at multiple time points and considering improvement at any one time point as a positive response. To assess the response following the CSF TT, 83% of respondents examined the patients at multiple time points, and 58% felt that if it were positive at any one time point, they would consider it positive. Although 66.1% would not refer patients for surgery following a negative TT, one‐third still considered surgical referral if imaging findings were strongly suggestive of NPH. MRI CSF flow studies were routinely used by 44% of respondents, although nearly half considered them primarily prognostic rather than decisive for surgical referral. The practice trends of surveyed physicians are summarized in Table 1.

4. DISCUSSION

Although NPH is more common than previously thought, nearly three‐fourths of practitioners evaluate fewer than two patients with NPH per month. From clinical diagnosis to surgical referral, we found wide variations in practice. Our findings suggest a lack of standardized criteria for determining CSF TT responsiveness and highlight important differences in real‐world evaluation practices across centers. Although the Japanese guidelines ¹ recommended evaluation when more than one symptom (gait, urinary, or cognitive) was present, nearly 30% of practitioners in our survey performed a CSF TT if at least one symptom was present. This shows the low threshold for evaluating patients by the respondents, probably because NPH is considered a treatable and under‐recognized cause of cognitive and gait impairment.

This is similar to the Relkin guidelines, which suggest that if imaging features are present with gait disturbances, a CSF tap test may be performed. ⁸ Similarly, most respondents performed assessments at multiple time points after the CSF TT, and more than half considered improvement at any one time point sufficient to define responsiveness. Such an approach may increase the sensitivity of the CSF TT and potentially increase the number of patients considered eligible for surgical intervention. This likely reflects the prevailing clinical perception that NPH is an under‐recognized yet potentially treatable condition in which missing a surgically responsive patient may have significant consequences. However, this could make it difficult to compare the outcomes across these centers due to differences in surgical referral patterns and baseline characteristics.

The CSF TT is considered a valuable measure for predicting a favorable response to shunt surgery in patients with iNPH. However, the volume of CSF tapped and the timing of assessment vary between practitioners, as seen in our data. Even in Japan, where a positive response is not considered a prerequisite for subjecting patients to surgery, almost all patients enrolled in clinical studies had undergone a CSF TT. ⁹

In line with global practice, 95% of practitioners performed a CSF TT before subjecting patients to surgery. However, only 66% of respondents said they would not refer the patient for a ventriculoperitoneal (VP) shunt if the TT were negative. Of interest, 36% of respondents used subjective improvement as a measure after the CSF TT, rather than scales, instrumental gait analysis, or cognitive batteries, to assess changes following a large‐volume CSF tap. This is again a source of significant heterogeneity and may have implications for the prognostication of post‐surgical outcomes due to misclassification in diagnosis.

Beyond the decision to perform the CSF TT itself, our survey also demonstrated marked heterogeneity in the methods used to define tap responsiveness. Respondents used a wide range of gait and cognitive assessment tools, including the Boon's gait scale, Timed up and go test (TUG) test, Mini‐Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), and the iNPH grading scale, with no single dominant approach. Even among practitioners using objective scales, there was substantial variability in the thresholds considered clinically meaningful. For example, respondents used differing MoCA/MMSE cutoffs ranging from 1‐ to 5‐point improvement, whereas both percentage‐based and absolute cutoffs were used for the TUG test. Of note a considerable proportion of respondents relied solely on subjective patient or clinician impression rather than standardized scales. Such variability may significantly influence the sensitivity of the CSF TT, affect reproducibility across centers, and contribute to differences in surgical referral patterns and reported postoperative outcomes.

According to a recent systematic review on this topic, there was wide variation in the timing of assessment, ranging from less than 3 h to more than 96 h after the CSF TT. This systematic review found that 70% of studies reported single‐point assessments, whereas the rest conducted multiple clinical assessments to evaluate symptom improvement. ⁵ In another systematic review of NPH gait assessment using instrumented gait measures, it was found that the variation in stable velocity was maximum around 24–48 h after CSF TT and progressively returned to baseline in 90–100 h. ¹⁰ In our survey, this variability was captured, and we found practitioners tapping different volumes of CSF, using various tests and cutoffs for assessing responses, as shown in the Table 1. In a recent cohort study from India, similar trends were captured. ¹¹

The literature is limited regarding the role of the CSF TT using the gait scale, the iNPH grading scale, and the MoCA/MMSE score in predicting shunt outcomes. In our recent meta‐analysis of the measures of prognostic accuracy for the CSF TT and lumbar infusion test, we found that the pooled specificity of the CSF TT was 52% (95% confidence interval [CI] 36–68), and the specificity of the lumbar infusion test was 67% (95% CI 50%–82%) (unpublished data). These studies have employed various definitions to assess CSF TT and shunt responsiveness, leading to considerable clinical heterogeneity. ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹

Our survey results reflect conflicting opinions regarding the predictive power and validity of the CSF TT, as up to a third of practitioners did not use the CSF TT to select patients for surgery if other tests, such as imaging and MRI flow studies, were suggestive of NPH. A recent systematic review on the diagnostic accuracy of the CSF TT also reported that the included studies used varying cutoffs to define both tap responsiveness and shunt responsiveness, highlighting the lack of consensus in this area. ²⁰ We need to conduct multi‐center studies to compare the various scales, determine cutoffs, and validate them. Meanwhile, we should develop consensus guidelines and definitions and encourage standardization, as has been done in the recent systematic review by Bluett et al. ⁵ This will enable the subsequent development of modules to train residents and practitioners, ensuring a uniform standard of practice. This would ensure that further studies may have homogeneous inclusion criteria to develop new markers of shunt responsiveness. In fact, because shunt surgery is an invasive procedure, it is essential to have sensitive and specific tests to predict shunt responsiveness.

This study has a few limitations worth noting. The study was subject to selection bias due to the lack of random sampling and self‐selection bias, as is often the case in most surveys.

The survey was disseminated via social media platforms, which may affect the representativeness of the sample and limit generalizability to the broader population of neurologists, especially in smaller settings. In this context, WhatsApp proved to be a practical tool for dissemination, offering a more reliable contact than email, which is often outdated or inactive. We received only 59 responses, which is a small number, with the majority of neurologists practicing in the tertiary care setting, and may not be truly representative of the entire community of neurologists. In addition, as we could not determine precisely the survey's reach, the response rate remains an estimate. Although demographic data were collected, the absence of random sampling means the findings cannot be assumed to reflect trends across all practitioners with similar backgrounds. We do not have information on the survey's geographic reach or the respondents' years of practice. We did not officially approach centers that perform regular NPH surgeries, and we do not have prospective information on whether postoperative outcomes matched clinical expectations based on their pre‐surgical evaluation. The study was also subject to information bias as the responses relied on the practitioners' memory and were not collected prospectively. The questionnaire focused solely on VP shunt as an option and did not discuss other therapeutic surgical procedures, such as endoscopic third ventriculostomy. Finally, we did not approach neurosurgeons, who, in some centers, are the sole health care providers involved in NPH management.

The lack of an interviewer also meant we could not verify participants' understanding of the questions, as evidenced by a couple of discrepancies in Table 1 reported by the respondents. Nonetheless, online surveys are increasingly recognized for their efficiency and accessibility, particularly in professional communities.

5. CONCLUSION

Overall, this survey offers valuable preliminary insights into current clinical practices and highlights areas for further structured research. NPH seems to be rarely seen by neurologists in tertiary care centers who evaluate fewer than two patients a month. Almost half of the physicians had fewer than five patients undergoing surgery in a year. There is widespread variability in the technique of the CSF TT and the post‐tap evaluation practices. Several physicians, however, seemed to rely on a subjective patient assessment based on visual evaluation rather than cutoffs or percentage improvement in scores. A consortium of experts evaluating the reliability and feasibility of these procedures can help in formulating guidelines and training modules for India, ensuring uniform practice.

CONFLICT OF INTEREST STATEMENT

None of the authors report any conflicts of interest. Author disclosures are available in the Supporting Information.

ETHICS STATEMENT

Ethics: The study was approved by the Institute Ethics Committee of the All India Institute of Medical Sciences, New Delhi, India vide AIIMSA2056/22.8.2024.

CONSENT STATEMENT

All participants provided informed written electronic consent.

Supporting information

Supporting Information: alz71584‐sup‐0001‐ICMJE.pdf

Elavarasi A, Poudel S, Dash D, et al. Current practice patterns among Indian neurologists in the Evaluation and management of Normal Pressure Hydrocephalus: A nationwide cross‐sectional survey. Alzheimer's Dement. 2026;22:e71584. 10.1002/alz.71584

DATA AVAILABILITY STATEMENT

All data generated have been included in the manuscript.