This article presents an efficient method for computing approximate confidence levels for searches for new particles when the expected signal and background levels are small, requiring Poisson statistics. The method allows combining results from multiple independent searches for the same particle, regardless of the variables used to discriminate between signal and background. Systematic uncertainties in the signal and background models are incorporated into the confidence levels. The procedure enables efficient computation of expected confidence levels. The method uses a likelihood ratio test statistic, which is the product of individual test statistics from each search channel. The confidence level for excluding the presence of a signal is calculated as the probability that the test statistic would be less than or equal to the observed value under the hypothesis of simultaneous signal and background. This probability is computed as a sum of Poisson probabilities. The Modified Frequentist confidence level is defined as the ratio of the confidence level for the signal plus background hypothesis to the confidence level for the background hypothesis. The method involves binning search results in discriminant variables and treating each bin as a separate search channel. This allows combining results from different analyses. The confidence levels are computed by summing probabilities of outcomes with test statistics less than or equal to the observed one. The method also accounts for systematic uncertainties in signal and background estimates by averaging over possible values of these estimates. The method is tested with numerical examples, including a simulation of a Higgs boson search. The results show that the method provides confidence levels close to those obtained from direct summations or Monte Carlo simulations. The method is also applied to combine results from multiple experiments, demonstrating its effectiveness in handling complex scenarios. The method has limitations, including the finite resolution of the test statistic PDF and the lack of incorporation of correlations between systematic uncertainties in different search channels. The technique is suitable for forming confidence limits near traditional levels like 90%, 95%, and 99%. It is also useful for efficiently scanning many possible models for signal production with different signatures. The method incorporates uncorrelated systematic uncertainties naturally and suggests Monte Carlo alternatives when correlated uncertainties are expected to be significant.This article presents an efficient method for computing approximate confidence levels for searches for new particles when the expected signal and background levels are small, requiring Poisson statistics. The method allows combining results from multiple independent searches for the same particle, regardless of the variables used to discriminate between signal and background. Systematic uncertainties in the signal and background models are incorporated into the confidence levels. The procedure enables efficient computation of expected confidence levels. The method uses a likelihood ratio test statistic, which is the product of individual test statistics from each search channel. The confidence level for excluding the presence of a signal is calculated as the probability that the test statistic would be less than or equal to the observed value under the hypothesis of simultaneous signal and background. This probability is computed as a sum of Poisson probabilities. The Modified Frequentist confidence level is defined as the ratio of the confidence level for the signal plus background hypothesis to the confidence level for the background hypothesis. The method involves binning search results in discriminant variables and treating each bin as a separate search channel. This allows combining results from different analyses. The confidence levels are computed by summing probabilities of outcomes with test statistics less than or equal to the observed one. The method also accounts for systematic uncertainties in signal and background estimates by averaging over possible values of these estimates. The method is tested with numerical examples, including a simulation of a Higgs boson search. The results show that the method provides confidence levels close to those obtained from direct summations or Monte Carlo simulations. The method is also applied to combine results from multiple experiments, demonstrating its effectiveness in handling complex scenarios. The method has limitations, including the finite resolution of the test statistic PDF and the lack of incorporation of correlations between systematic uncertainties in different search channels. The technique is suitable for forming confidence limits near traditional levels like 90%, 95%, and 99%. It is also useful for efficiently scanning many possible models for signal production with different signatures. The method incorporates uncorrelated systematic uncertainties naturally and suggests Monte Carlo alternatives when correlated uncertainties are expected to be significant.