This is one of my own pet medical peeves, stimulated recently by a thread here on Lyme tests and what bands meet criteria, etc. It relates directly to the difficulties with negative tests for Chlamydia pneumoniae. Yes, I know it's a complex issue, but the basic rule is really a simple one:
A serology test, either negative or positive, does not a diagnosis make!
Medical diagnosis is a complex art which includes serology, clinical history, symptom patterns, treatment history, and other medical tests. True, some serology carries more weight in the clinical equation that others and may be taken almost defacto as diagnosis. For example, one is treated for many diseases on the basis of positive serology, such as for TB or STD's, even if you are asymptomatic and have nothing obvious in your history.
But- the same does not hold true with negative tests, or with tests like Lyme's where the strict CDC criteria (specific bands must be positive) are not met. There are many reasons for negative tests and for missing criterian bands to occur even with the presence of the actual disease organism in the body. This is the complex part I won't go into here. It's really discussed more fully elsewhere on www.Cpnhelp.org
With a negative serology test or lack of full criterian bands, other things are taken into account in good medical diagnosis: history, clinical picture, and even empirical response to an empirical treatment is legitimate diagnostic information. Criterian which designed to meet the needs of a research study, in order to weed out false positives and thus have a more homogeneous test group, are not the same requirements of a medical practice, where a doctor is obligated to make a best attempt to help the patient, even if not everything is crystal clear on serology tests.
Of course, we do sacrifice certainty when we receive treatment this way. Someone, for example, who has negative serology for Cpn, but suffers a Cpn related disease, and who uses a CAP and has reactions or improvements, may have these from Cpn or from some other organisms (such as borrealis). And vice versa. Some people not meeting criteria for Lyme who treat with antibiotics and flagyl anyway, could be treating Cpn. The famous Lyme herx/die-off is almost indistinguishable from the famous Cpn herx/die-off. It's likely that we are killing more than one organism off anyway, especially in diseases which produce immunocompromised individuals, like Cpn.
Now, there are other signs which tend to confirm one organism or the other. For example, as far as we know only Cpn creates secondary porphyria because of it's stealing of host ATP. These porphyric symptoms are distinct in many ways from endotoxin or cytokine symptoms, common when killing a number of organisms. So we can get empirical hints from empirical treatment, which may help clarify an empirical diagnosis.
What we may not have is certainty. But most of us can sacrifice a little certainty for getting better. I'd rather be well and uncertain, than still sick and looking for absolute guarantee!