Diagnosis and treatment patterns among patients with newly diagnosed Helicobacter pylori infection in the United States 2016–2019
This is the first population-based analysis in the US assessing current HP treatment patterns and eradication rates in the US based on real-world data. In this analysis of US claims and EHR data, 45.3% of patients with HP infection received either no guideline recommended treatment or a potentially inappropriate regimen based on prior antibiotic use. Clarithromycin triple therapy was the most commonly prescribed first-line and second-line regimen despite macrolide exposure in the baseline period for more than 20% of patients, and first-line use of clarithromycin triple therapy was observed in nearly half of patients receiving second-line clarithromycin. triple therapy. Approximately one-third of patients did not undergo post-treatment HP eradication testing, and of the subgroup of patients with available data, 20% had no first-line HP eradication treatment. These real-world findings suggest that there is potential to significantly improve HP treatment and management gaps consistent with current evidence-based guidelines.
The most recent ACG guidelines, published in 2017, recommend against the use of clarithromycin triple therapy as a first-line treatment. except for specific criteria are met: (1) regional resistance rates to clarithromycin are less than 15%, (2) the patient does not have costly macrolide exposure, and (3) the specific strain is sensitive to clarithromycin13. In the current study, more than 80% of patients received triple clarithromycin therapy as first-line therapy and more than half received it as second-line treatment, which is consistent with previous smaller studies but the first to demonstrate it in a general US population19,20,21. Our findings suggest that the first-line regimen used was often inappropriate based on patients’ previous exposure to antibiotics. For example, 21.0% of patients receiving a first-line regimen of clarithromycin had received a prescription for a macrolide antibiotic in the 12-month pre-index period alone. Particularly because HP is usually acquired during childhood5, the percentage of patients who received a macrolide prescription after infection but prior to first-line HP eradication therapy is likely even higher. Our findings are consistent with a previous analysis of the PharMetrics Plus database, which found that 18% of patients receiving a clarithromycin-based first-line regimen had received a prescription for macrolides in the previous year19. We saw similar trends with first-line treatments containing levofloxacin and prior exposure to fluoroquinolone.
Although local resistance rates are not widely available, recent US studies have consistently shown resistance to clarithromycin to be greater than 15%16,17,22,23. Nevertheless, our analysis showed that the clarithromycin prescribing pattern remained fairly stable over the study period, decreasing only slightly from 81.4% in 2016 to 78.1% in 2019. Note that most of our study period preceded the approval of a fixed-dose triple therapy with rifabutin and would not reflect the impact of this approval on treatment patterns. Despite being strongly recommended for use as a first-line therapy, only 6.6% of patients received bismuth quadruple therapy in the first-line setting. In addition, the choice of second-line therapy was often inappropriate as approximately half (50%) of patients receiving second-line therapy received the same treatment as during first-line treatment, contrary to recommended guidelines. Contrary to guideline recommendations, more than 90% of patients receiving the same first-line and second-line treatment received two cycles of clarithromycin triple therapy15.
The recently updated Maastricht V/Florence Consensus Report suggested susceptibility testing prior to prescribing first-line treatment to align with antibiotic management practices; despite the recommendation, they note that routine use of susceptibility testing in clinical practice has yet to be established and may not always be feasible14. In the absence of susceptibility testing, the treating physician should use local antibiotic resistance patterns to guide decision making regarding the use of clarithromycin triple therapy as first-line treatment13.14. A knowledge gap remains in this due to the lack of a US national registry of local and regional resistance patterns, which could explain the continued heavy reliance on clarithromycin triple therapy.
In addition to gaps between recommended eradication treatment and actual treatment patterns, this study also identified gaps between recommended and field-test practices for post-eradication treatment. As noted earlier, persistent HP infection puts patients at constant risk for complications associated with chronic HP infection. In light of rising antibiotic resistance rates and rising eradication failure rates, it is critical that patients undergo repeated non-serologic testing, ideally at least 4 weeks after completion of eradication treatment, to confirm eradication. In addition, resolution (and similarly persistence) of symptoms is not correlated with success versus failure of eradication treatment and thus is not a substitute for non-serologic testing after eradication treatment. In this claims-based study, only 67.4% of patients underwent a diagnostic HP test after first-line treatment to confirm successful eradication. Although this is higher than previously reported percentages, which ranged from 24 to 44%24.25it still suggests that one-third of HP patients complete eradication therapy without evaluating whether that eradication was successful. Further research is needed as to whether this is due to a missed opportunity by healthcare providers to recommend testing or a missed opportunity by patients to return for physician-recommended testing.
In the subset of patients with post-treatment non-serologic HP test results available, more than 20% had persistent HP infection after first-line treatment, including 20% of patients receiving triple clarithromycin therapy and 15% of those receiving bismuth quadruple therapy. These findings are consistent with other real-world US studies, including one of the 1966 US military that reported eradication failure in 25% of all patients, 24% failure in patients receiving triple clarithromycin therapy, and 15% failure in patients receiving bismuth. quadruple therapy26. Similarly, a retrospective analysis of 1101 patients treated in Rhode Island reported that eradication failed in 23% of patients receiving triple clarithromycin therapy and 14% of patients receiving quadruple bismuth therapy27.
Our data suggest that not all patients who failed HP eradication treatment received subsequent treatment. Given the limited number of post-eradication confirmation tests, we were limited in our ability to discern the eradication failure rate for the entire cohort. That said, only 11% of patients who received first-line therapy progressed to second-line therapy. Since an overall eradication rate of 89% is unlikely given the findings of the subgroup analysis and previous studies, this suggests that many patients may have persistent HP infection. Possible reasons for this discrepancy include that some healthcare providers may not offer treatment if patients report that symptoms have resolved or that patients who experience a resolution or significant improvement in symptoms are less likely to complete another course, especially given the fact that side effects are common with HP treatment. As noted above, resolution of symptoms does not equate to eradication of HP and patients remain at risk for further complications with persistent HP infection.
This retrospective analysis used routinely collected claims and EHR data and is subject to the typical limitations of using retrospective data. The data set used in this analysis included insured individuals, which may additionally affect generalizability to the entire US population, as approximately 15% of US adults aged 18 to 64 were uninsured in 201928. These findings also may not be generalizable to individuals with certain types of insurance not captured in this data set, such as those who receive their health care through the Veterans Health Administration. In addition, only a small subset of patients had non-serologic laboratory results following HP eradication treatment, which may impact the accuracy of our estimated eradication rate, as well as generalizability, as patients who had repeat tests ordered and completed may have a higher risk group compared to the group of patients who did not.
We used a strict definition of guideline-recommended regimens, so we may have missed patients who received alternative regimens outside the guidelines, but still with the intention of treating HP. We cannot distinguish between concurrent, sequential, and hybrid therapy using claims data, so we combined these regimens; although we cannot draw separate conclusions for these treatments, they comprised a minority of regimens: 5.1% of first-line and 5.8% of second-line regimens. As with any retrospective data analysis, we cannot confirm patient adherence even if the full prescribed course has been filled.
We were unable to determine the exact date of the HP acquisition, which is a common limitation that is difficult to overcome. Previous studies in endemic populations suggest that HP acquisition is most common in early childhood, which is supported by limited US data, at least in families with an infected parent29. In addition, we cannot confirm details regarding the diagnostic EGD, including whether or not gastric biopsies were obtained and, if so, their number and location; whether intraprocedure HP testing has been performed; nor can we confirm whether special pathology stains have been used for HP detection on gastric biopsies. Similarly, the use of HP antibody sensitivity testing was very rare, and we cannot comment on the antibody sensitivity of HP strains.
Finally, our follow-up time, while sufficient to determine treatment and follow-up practice patterns, is insufficient to thoroughly evaluate the downstream consequences of eradication failure. This is an important area of research in light of rising eradication failure rates. Future research should investigate whether the implementation of sequence sets or other clinical decision support tools can increase adherence to guidelines and improve patient outcomes.