CMS’s Claims Processing Manual (CPM) is a rich resource for understanding how Medicare claims are submitted for reimbursement, and thus, a viable resource for identifying potential methodologies for research. This article seeks to quantify how often clinical trials can be found using information from the CPM. Applying submission requirements to 2019 Medicare FFS Inpatient, Outpatient, and Carrier claims data resulted in a modest number of claims and beneficiaries.
Can we use the Claims Processing Manual to define service utilization?
The Claims Processing Manual (CPM) is the rulebook for submitting claims to CMS. It covers a wide range of medical coding instructions by clinical setting and form. As the main source of instruction on claims processing and assuming that providers are incentivized to follow submission requirements for reimbursement, it should be a reliable source for defining research methodology.
We tested this on clinical trial identification in Medicare Fee-for-Service (FFS) claims data. Clinical trials are a type of research study where individuals are assigned to different interventions to evaluate outcomes. They are essential for identifying new forms of treatment and better understanding existing treatment. Clinical trials were chosen because routine costs are covered by Medicare and the CPM offers clear and precise instructions on claims submission.
Clinical Trial Identification and the Claims Processing Manual
Tables 1 and 2 provide a list of CPM requirements for clinical trial claims(1). This is a high-level review and doesn’t include coding requirements for unique situations. The tables are split by institutional and non-institutional claims. Institutional claims serve inpatient, outpatient, hospice, skilled nursing facilities, and home health care facilities. Non-institutional claims serve individual providers (also known as carrier) and durable medical equipment claims.
Table 1. General Billing Requirements for Institutional Claims
| Variable | Value | Description |
|---|---|---|
| Claim Diagnosis Code | ICD-9: V70.7 ICD-10: Z00.6 | This code must be in the primary or secondary position. |
| HCPCS Modifier Code | Q0, Q1 | Only for OP claims |
| Claim Related Condition Code | 30 | |
| Claim Value Code | D4 | |
| Claim Value Amount | Clinical trial number, 99999999: Unknown |
Table 1 caption: This table describes the institutional billing requirements for clinical trial claims.
Table 2. General Billing Requirements for Non-Institutional Claims
| Variable | Value | Description |
|---|---|---|
| Claim Diagnosis Code | ICD-9: V70.7 ICD-10: Z00.6 | This code must be in the primary or secondary position. |
| Clinical Trial Number | Clinical Trial Number 99999999: Unknown | |
| HCPCS Modifier Code | Q0, Q1 |
Table 2 caption: This table describes the non-institutional billing requirements for clinical trial claims.
Methodology
The criteria identified in Tables 1 and 2 were applied to 2019 5% Carrier, 5% Outpatient, and 100% Inpatient FFS claims data. No cleaning or reduction was applied to these data. 5% Carrier and 5% Outpatient counts were multiplied by 20 to simulate 100% files.
For Inpatient and Outpatient claims, clinical trial claims are defined as having, at least, one of the flags listed in Table 1. For Carrier claims, clinical trial claims are defined as having, at least, one of the flags listed in Table 2. Clinical Trial Numbers are not considered flags for this analysis, and the Claim Diagnosis Code inclusion flag was broadened to allow for Z00.6 in any position.
Null clinical trial numbers are defined as unknown or missing values (Blank, 0-filled, 9-filled, 0, 999999.99, or 9999999). Null values were taken from the CPM and manually identified.
Results
Beneficiaries and claims can be identified using the information provided in the CPM. Using this methodology, clinical trial claims and associated beneficiaries are a relatively small proportion. Between 0.2% and 0.9% of all claims have some type of clinical trial identifier (Table 3). The beneficiaries associated with these claims are approximately 0.33% and 1.9% of the population by setting (Table 4).
Interestingly, although the CPM instructs medical coders to list Z00.6 as the primary or secondary diagnosis, this occurs between 52.3% and 81.8% of the time. In a similar vein, although reporting clinical trials has been mandated since 2014(2), between 1.0% and 2.8% of clinical trial claims contain unknown or missing clinical trial numbers.
Table 3. Clinical Trial Claim Counts by Identification Criteria and FFS File, 2019
| Carrier | Inpatient | Outpatient | |
|---|---|---|---|
| Claims with Clinical Trial Diagnosis Codes | 929,940 | 89,870 | 411,240 |
| Claims with Clinical Trial Diagnosis Codes in Primary or Secondary Positions | 760,940 | 46,977 | 322,580 |
| Claims with Clinical Trial Modifier Codes | 824,820 | 23 | 410,060 |
| Claims with Clinical Trial Value Codes | NA | 88,789 | 421,280 |
| Claims with Clinical Trial Condition Codes | NA | 89,866 | 411,240 |
| Null Clinical Trial Number Count | 54,340 | 915 | 12,340 |
| Clinical Trial Claim Count | 2,495,900 | 91,373 | 436,020 |
| Total Number of Claims | 942,061,760 | 10,670,564 | 180,329,040 |
Table 3 caption: This table describes the total number of clinical trial claims based on various identification methods. The bottom row contains the total number of claims by file.
Table 4. Count of Beneficiaries Associated with Clinical Trials, 2019
| Carrier | Inpatient | Outpatient | |
|---|---|---|---|
| Clinical Trial Beneficiary Count | 719,840 | 87,912 | 87,940 |
| Total Number of Beneficiaries | 37,068,260 | 6,273,111 | 26,604,040 |
Table 4 caption: This table describes the total number of beneficiaries associated with clinical trial claims by file. The bottom row contains the total number of beneficiaries by file.
Discussion
Using the information from the CPM results in a modest number of clinical trial claims and beneficiaries. Depending on the file, clinical trial claims represented around 0.2% and 0.9% of claims and clinical trial beneficiaries represented around 0.33% and 1.9% of all beneficiaries. Without the ability to pull medical history and patient information from every clinical trial, it will be difficult to ascertain a gold standard that can be applied across CMS FFS data.
There are studies that look at completeness at a specific clinical trial level. One such study found that around 75% of their cohort had, at least, one claim with a correctly reported clinical trial number(3). Thus, our results could indicate low enrollment of Medicare beneficiaries in clinical trials. It could also indicate imperfect or conflicting claims submissions by providers. For example, Z00.6 not being in the first or second position could indicate administrative error or management of conflicting claims processing instructions. Ultimately, the instructions in the CPM can be used to identify claims and beneficiaries, but when limited to CMS claims data, it is difficult to determine if claims and beneficiaries are missing.
Limitations
The CPM only instructs how providers should submit claims for reimbursement. Although the likelihood of compliance is higher because it is related to payment, it does not guarantee that all claims will be submitted exactly as mentioned in the CPM.
Without having a gold standard for comparison, it is difficult to determine the completeness of clinical trial claim and beneficiary data.
For institutional claims, clinical trial numbers had accounting formats applied. This made it difficult to identify null values, so the null value list may not be all encompassing. This may result in an undercount of null clinical trial numbers.
Conclusion
Although it will require a gold standard comparison to confirm utility, the CPM can still provide a solid starting point for defining service utilization in CMS FFS data. As seen with clinical trials, it can provide clear, definitive instructions that easily apply to data analysis. It can also inform researchers about background and potential caveats. It is a thorough and helpful resource for researchers interested in CMS FFS data.
1. Medicare Claims Processing Manual [Internet]. Centers for Medicare & Medicaid Services; [cited 2026 Jan 22]. Available from: https://www.cms.gov/regulations-and-guidance/guidance/manuals/downloads/clm104c32.pdf
2. Brocato-Simons P, Hakim R. Mandatory Reporting of National Clinical Trial (NCT) Identifier Numbers on Medicare Claims - Qs & As [Internet]. Centers for Medicare & Medicaid Services; 2014 Oct [cited 2025 Dec 23]. Available from: https://www.cms.gov/medicare/coverage/coverage-with-evidence-development/downloads/mandatory-clinical-trial-identifier-number-qsas.pdf
3. Green AK, Tabatabai SM, Bai X, Mishra Meza A, Lesny AM, Aghajanian C, et al. Validation of a Population-Based Data Source to Examine National Cancer Clinical Trial Participation. JAMA Netw Open. 2022 Mar 22;5(3):e223687.
4. Hammill BG, Hernandez AF, Peterson ED, Fonarow GC, Schulman KA, Curtis LH. Linking inpatient clinical registry data to Medicare claims data using indirect identifiers. Am Heart J. 2009 Jun 1;157(6):995–1000.
5. Hlatky MA, Ray RM, Burwen DR, Margolis KL, Johnson KC, Kucharska-Newton A, et al. Use of Medicare Data to Identify Coronary Heart Disease Outcomes in the Women’s Health Initiative. Circ Cardiovasc Qual Outcomes. 2014 Jan;7(1):157–62.