claim-raven

We Classified 16,800 BVA Denials, and here are the top 5 reasons.

When a claim gets denied at the Board of Veterans' Appeals, the decision letter gives a reason. Sometimes it's straightforward. Sometimes it reads like the Board is describing a completely different case than the one you filed. But across thousands of decisions, patterns emerge — and those patterns are worth understanding.

I looked at 101,518 BVA decisions across 185 conditions and classified the primary denial reason for every denied case where the Board's language made the reason identifiable. Out of roughly 41,400 denials, we were able to classify a specific primary reason in about 16,800 cases. What I found was that five denial categories account for nearly 90% of those classified denials — and each one correlates with distinct evidence patterns and nexus quality profiles. The breakdown isn't random. There's a data story behind each one, and some of it honestly surprised me.

---

TL;DR

Severity Insufficient is the #1 denial reason at 4,979 classified denied cases (29.1%) — and it's not always about how sick you are.
Nexus Gap follows at 3,913 cases (22.9%) — the connection between service and condition wasn't established to the Board's satisfaction.
Service Connection Missing accounts for 2,911 cases (17.0%) — the broader evidentiary picture didn't establish a link to service.
Diagnosis Missing hits 2,793 cases (16.3%) — no confirmed diagnosis, no claim.
Nexus quality is the clearest differentiator: cases rated "Strong" had an 89.5% grant rate versus 3.3% for "Weak" and 1.5% for "Missing."
These five reasons together cover 89.3% of classified denials in the dataset.

---

The Big Picture: Where Denials Cluster

Before diving into each denial reason, here's the full landscape. Out of 101,518 total cases, the overall denial rate was 40.8%, with a 29.5% grant rate and 26.6% remand rate. That means roughly 4 in 10 cases that make it all the way to the BVA end in a denial.

Not every denied case had a cleanly classifiable denial reason — many BVA decisions cite overlapping factors or use language that doesn't map neatly to a single category. Of the roughly 41,400 denied cases, we classified a specific primary denial reason for about 16,800. The remaining cases had denial reasons that were either too case-specific to categorize, cited multiple overlapping factors, or used language our classification system couldn't cleanly parse.

Here's how those classified denials break down by reason:

Severity Insufficient: 4,979 cases (29.1%)
Nexus Gap: 3,913 cases (22.9%)
Service Connection Missing: 2,911 cases (17.0%)
Diagnosis Missing: 2,793 cases (16.3%)
Procedural Issue: 685 cases (4.0%)
Timeliness Issue: 684 cases (4.0%)
In Service Event Missing: 613 cases (3.6%)
Presumption Not Met: 103 cases (0.6%)
Evidence Not Credible: 57 cases (0.3%)
Stressor Not Verified: 57 cases (0.3%)

The top three alone — Severity Insufficient, Nexus Gap, and Service Connection Missing — make up 69.0% of classified denials. That's a pretty tight cluster, and it tells you where the friction points are in the system.

What's interesting is that these aren't evenly distributed across conditions. Some conditions get hammered by one denial type far more than others. And the nexus quality ratings in denied cases tell a consistent story about what the Board was looking at when it said no.

---

Denial Reason #1: Severity Insufficient (4,979 Cases — 29.1%)

This is the big one, and it's probably the most misunderstood denial reason. "Severity Insufficient" doesn't always mean the Board thinks you're fine. It often means the evidence in the record didn't demonstrate that the condition meets the rating criteria for the level claimed — or in some cases, for any compensable rating at all.

When I looked at the nexus quality breakdown across the full dataset, the pattern was stark. Cases with Strong nexus quality had an 89.5% grant rate across 12,509 cases. Cases rated Weak dropped to 3.3% across 18,260 cases. And cases where nexus was Missing entirely sat at 1.5% across 10,718 cases.

Now, severity denials are a slightly different animal because they're often about the degree of disability rather than whether service connection exists at all. But the nexus quality data still matters here because the Board frequently weighs medical opinions when deciding severity — particularly when there's a dispute between a VA examiner's findings and private medical evidence.

The conditions where severity denials cluster tend to be ones with subjective symptom reporting — mental health conditions, musculoskeletal issues, and conditions where the rating criteria hinge on functional impairment that can be measured different ways depending on the examiner. A veteran might report significant daily limitations while a C&P exam notes "mild" findings. That gap between lived experience and examination results shows up repeatedly in severity denial decisions.

I don't know exactly why this is the single largest denial category by such a wide margin, but I have a few theories. Rating criteria for many conditions are built around clinical findings that may not capture the veteran's actual functional impairment. And the Board often defers to the most recent C&P exam even when other medical evidence paints a different picture — especially if the other evidence isn't structured around the specific rating criteria.

The pattern I noticed: severity denials tend to correlate with cases where the medical evidence exists but doesn't speak directly to the rating schedule's specific benchmarks.

---

Denial Reason #2: Nexus Gap (3,913 Cases — 22.9%)

This one is more straightforward in concept but brutal in practice. A "Nexus Gap" denial means the Board found that the evidence didn't adequately link the veteran's current condition to their military service.

And this is where the nexus quality data really screams. The spread across the dataset is enormous:

Strong nexus: 89.5% grant rate (12,509 cases)
Adequate nexus: 68.9% grant rate (18,597 cases)
Weak nexus: 3.3% grant rate (18,260 cases)
Missing nexus: 1.5% grant rate (10,718 cases)
Inadequate nexus: 0.0% grant rate (149 cases)

That gap between Adequate (68.9%) and Weak (3.3%) is probably the most striking number in this entire analysis. It's not a gradual decline — it's a cliff. Cases that crossed some threshold of nexus quality had dramatically different outcomes, and the drop-off is severe.

I'm not sure what to make of the sharpness of that divide, honestly. One possibility is that "Weak" and "Missing" nexus cases are ones where the veteran essentially had no medical opinion connecting service to condition — maybe just a diagnosis and service records, with the Board expected to connect the dots on its own. Another possibility is that the nexus opinions in those cases existed but were so conclusory that the Board gave them no weight.

Either way, the data pattern is consistent: the nexus opinion isn't just a factor in BVA decisions — it appears to be the factor in a huge number of cases. The difference between a 68.9% grant rate and a 3.3% grant rate isn't a marginal difference. It's a different universe of outcomes.

When you cross-reference nexus gap denials with connection types, secondary claims are particularly interesting. Secondary service connection requires proving that an already service-connected condition caused or aggravated the new one — which is a nexus argument by definition. Some secondary chains in the data, like Anxiety Secondary Tinnitus at a 4.9% grant rate (865 cases) or ED Secondary PTSD at 5.6% (499 cases), show very low success rates. Meanwhile, conditions with clearer established medical pathways tend to perform better. That pattern suggests the strength of the underlying medical rationale matters a lot for secondary connections.

---

Denial Reason #3: Service Connection Missing (2,911 Cases — 17.0%)

At first glance, this looks like it might be the same thing as Nexus Gap. But there's an important distinction in how the Board uses these terms.

"Service Connection Missing" typically appears when the Board finds that the overall evidentiary picture doesn't establish a connection to military service — not just that the nexus opinion was weak, but that the broader case for service connection wasn't made. This can include situations where:

There's no evidence of an in-service event, injury, or illness that could plausibly connect to the current condition
The service treatment records are silent on anything related to the claimed condition
The timeline between service and the current diagnosis is long, with no bridging evidence

This denial reason accounted for 2,911 cases (17.0% of classified denials) — the third most common category.

When I cross-reference this with the lowest grant rate conditions in the dataset, a pattern emerges. Conditions like Stroke at just 1.0% (203 cases) and Heart Disease Secondary Sleep Apnea at 6.8% (500 cases) and Hypertension Secondary Sleep Apnea at 7.4% (665 cases) are conditions where establishing the service connection pathway is inherently difficult. The medical literature linking these conditions to service (or to other service-connected conditions) is either limited or contested, and that makes the evidence burden heavier.

Compare that to conditions like Anxiety at 49.7% (1,197 cases) where the connection to service — particularly to traumatic experiences — is more widely recognized in the medical community and more frequently supported by both VA and private examiners.

The data can't tell me whether the difference in grant rates is driven by the actual strength of the medical connection or by how willing medical providers are to write supportive opinions for certain conditions versus others. It could be both. But the spread is real, and it follows a pattern that correlates with how well-established the medical pathway is.

---

Denial Reason #4: Diagnosis Missing (2,793 Cases — 16.3%)

This one seems like it would be simple — either you have a diagnosis or you don't. But in practice, it's more complicated than that.

A "Diagnosis Missing" denial means the Board found that the veteran didn't have a current, confirmed diagnosis of the claimed condition at the time of the claim or during the appeal period. This comes up in a few distinct scenarios:

First, there are cases where the veteran has symptoms but no formal diagnosis in the record. This is common with conditions that are subjective or that overlap with other diagnoses — a veteran might experience chronic pain, fatigue, and cognitive issues that they attribute to one condition, but the medical record either doesn't include that specific diagnosis or attributes the symptoms to something else.

Second, there are cases where a diagnosis existed at some point but isn't current. If a condition resolved before or during the claim period, the Board may find that the diagnostic requirement isn't met — even if the veteran is still experiencing residual effects.

At 2,793 cases and 16.3% of classified denials, this is a significant chunk. And it overlaps with the nexus quality data in an important way: if there's no confirmed diagnosis, there's typically no nexus opinion either, because a nexus opinion by definition connects a diagnosed condition to service. So many of these cases likely also had Missing nexus quality — which, as the data shows, corresponds to just a 1.5% grant rate across 10,718 cases.

The conditions where this denial reason hits hardest tend to be ones where diagnosis is inherently complex or contested — conditions with overlapping symptom profiles, undiagnosed illness claims, and situations where VA examiners and private providers disagree on whether a diagnosis is warranted.

For context, the overall dataset includes some conditions with very high grant rates where diagnosis is rarely in question — Prostate Cancer at 49.9% (597 cases) and Parkinson's Secondary Agent Orange at 51.8% (500 cases) — suggesting that when the diagnosis is clear and well-documented, the Board spends less time on that threshold question and more time on the merits.

---

Denial Reason #5: Timeliness Issue (684 Cases — 4.0%)

The fifth most common denial reason drops off significantly in volume — 684 cases at 4.0% of classified denials — but it's worth noting because it's the one denial reason that has nothing to do with the strength of the evidence.

Timeliness denials happen when the Board finds that a claim or appeal wasn't filed within the required timeframe. This can involve the one-year window after a rating decision, the deadline for filing a Notice of Disagreement, or other procedural timelines.

What makes this category different from the others is that the merits of the case are essentially irrelevant. Even a veteran with Strong nexus quality and overwhelming evidence can lose on timeliness grounds. The Board generally doesn't have discretion to waive these deadlines unless specific equitable tolling criteria are met.

At 684 cases, this is a relatively small slice of the overall denial picture. But for the veterans in those cases, it's a particularly frustrating outcome — the denial isn't about whether the claim had merit, but about whether it was filed on time.

---

The Remaining Categories: Smaller Denial Reasons

The remaining denial reasons collectively account for about 8.5% of classified denials:

Procedural Issue: 685 cases (4.0%)
In Service Event Missing: 613 cases (3.6%)
Presumption Not Met: 103 cases (0.6%)
Stressor Not Verified: 57 cases (0.3%)
Evidence Not Credible: 57 cases (0.3%)

A couple of these are worth brief mention. In Service Event Missing (613 cases) is particularly relevant for conditions like PTSD where the stressor verification process — or the broader question of whether an in-service event occurred — is central to the claim. It's related to but distinct from Stressor Not Verified (57 cases), which specifically refers to PTSD stressor corroboration.

Evidence Not Credible at just 57 cases (0.3%) is notable for how rare it is. Despite how often veterans worry about being "caught" or not believed, the Board explicitly citing credibility as the denial reason is extremely uncommon in this dataset.

---

The Nexus Quality Thread That Runs Through Everything

If there's one data pattern that ties all of these denial reasons together, it's nexus quality. Here's the full breakdown one more time, because it bears repeating in context:

Strong: 89.5% grant rate across 12,509 cases
Adequate: 68.9% grant rate across 18,597 cases
Weak: 3.3% grant rate across 18,260 cases
Missing: 1.5% grant rate across 10,718 cases
Inadequate: 0.0% grant rate across 149 cases
Mixed: 0.0% grant rate across 5 cases

That cliff between Adequate and Weak — from 68.9% down to 3.3% — shows up everywhere. It correlates with every denial category because the Board's evaluation of nexus quality influences whether it finds severity established, whether it finds service connection established, and sometimes even whether it accepts a diagnosis.

Cases with Strong or Adequate nexus quality had grant rates of 89.5% and 68.9% respectively. Cases with anything below that — Weak, Missing, Inadequate — were functionally denied at the Board level, with grant rates of 3.3% or lower.

I don't know exactly why the drop-off is so sharp rather than gradual. It could be that "Adequate" represents a threshold — once a nexus opinion meets certain baseline criteria (addresses the right question, provides a rationale, is based on review of the record), the Board gives it meaningful weight. Below that threshold, it's as if the opinion barely exists.

The nexus quality data covers 60,238 cases where we could assess nexus quality from the decision text. The remaining cases either didn't involve nexus quality as a factor (such as pure rating disputes) or used language that didn't allow a clear quality assessment.

---

Bottom Line

Five denial reasons account for the vast majority of classified BVA denials in this dataset: Severity Insufficient (29.1%), Nexus Gap (22.9%), Service Connection Missing (17.0%), Diagnosis Missing (16.3%), and Timeliness (4.0%). Together, that's 89.3% of classified denials captured in five categories.

The through-line across almost all of them is nexus quality. The difference between cases with Strong nexus opinions and those with Weak ones isn't subtle — it's 89.5% versus 3.3%. That pattern holds across conditions, connection types, and denial categories. Whether the Board is evaluating severity, service connection, or even the relevance of a diagnosis, the quality of the medical opinion connecting everything together appears to be the single most important variable in the dataset.

---

Methodology And Limitations

Data source: Analysis of 101,518 BVA decisions across 185 conditions in the Claim Raven case database.
Denial categorization: Out of approximately 41,400 denied cases, we classified a specific primary denial reason for about 16,800. Denial reasons were classified from Board decision language into standardized categories. Many denied cases cited overlapping factors, used case-specific language, or couldn't be cleanly mapped to a single category — these were excluded from the denial reason breakdown. Categories are not mutually exclusive — some cases may have had multiple denial reasons, with the primary one recorded.
Nexus quality ratings: Nexus quality was assessable in 60,238 of the 101,518 total cases. Cases were rated on a scale from Missing to Strong based on how Board decisions described the medical opinion evidence. This is a classification system, not a direct quote from the Board — the ratings reflect our assessment of how the Board characterized the evidence. Cases not included in the nexus breakdown either didn't involve nexus as a factor or used language that didn't allow a clear quality assessment.
Grant rate calculations: Grant rates by nexus quality are computed across all cases where nexus quality was assessed, not just denied cases. They reflect the overall outcome distribution at each nexus quality level.
Limitations:
These are BVA-level cases only — they represent claims that were appealed, not the full universe of VA claims. Cases resolved favorably at lower levels aren't in this dataset.
Denial reasons were classifiable for roughly 40% of denied cases. The unclassified denials may have different patterns than the ones analyzed here, though there's no reason to believe the classified subset is unrepresentative.
Denial reasons may overlap — a case denied for "Nexus Gap" might also have had severity issues, but only the primary denial reason was captured.
Nexus quality is assessed from decision text, which introduces some subjectivity in classification.
Condition-level grant rates reflect cases that made it to appeal, which creates a selection bias — conditions with high initial grant rates may only show up at the BVA when they're unusually difficult cases.
Individual outcomes depend on facts specific to each case. Aggregate patterns are useful for understanding the landscape, not for predicting any single result.

---

Disclaimer

I'm not accredited by VA, not a lawyer, not a VSO. This is data analysis, not claim advice. These are patterns from cases that made it to the BVA - they don't predict individual outcomes. If you need help with your claim, work with an accredited representative.